Pneumonia Detection Challenge¶

Project Description¶

Now to detection Pneumonia we need to detect Inflammation of the lungs. In this project, you’re challenged to build an algorithm to detect a visual signal for pneumonia in medical images. Specifically, your algorithm needs to automatically locate lung opacities on chest radiographs.

Data Description¶

stage_2_train.csv - This is the training set. Contains patientIds and bounding box / target information.
stage_2_detailed_class_info.csv - provides detailed information about the type of positive or negative class for each image.
stage_2_train_images - Images for training. Total 26,684 dicom images.
stage_2_test_images - Images for testing. Total 3000 dicom images.
Data Fields
1. patientId - A patientId. Each patientId corresponds to a unique image.\n
2. x - the upper-left x coordinate of the bounding box.
3. y - the upper-left y coordinate of the bounding box.
4. width - the width of the bounding box.
5. height - the height of the bounding box.
6. Target - the binary Target, indicating whether this sample has evidence of pneumonia.
7. class - Normal, No Lung Opacity / Not Normal, Lung Opacity

!pip install pydicom #We are going to use pydicom package for dicom images processing.

Collecting pydicom
  Downloading https://files.pythonhosted.org/packages/d3/56/342e1f8ce5afe63bf65c23d0b2c1cd5a05600caad1c211c39725d3a4cc56/pydicom-2.0.0-py3-none-any.whl (35.4MB)
     |████████████████████████████████| 35.5MB 1.2MB/s 
Installing collected packages: pydicom
Successfully installed pydicom-2.0.0

#import necessory packages
import os
import csv
import random
import pandas as pd 
import numpy as np
import matplotlib
import matplotlib.pyplot as plt
from tqdm import tqdm_notebook
from matplotlib.patches import Rectangle
import seaborn as sns
import pydicom as dcm
from skimage import io
from skimage import measure
from skimage.transform import resize
import tensorflow as tf
from tensorflow import keras
import matplotlib.patches as patches
import pydicom
%matplotlib inline

/usr/local/lib/python3.6/dist-packages/statsmodels/tools/_testing.py:19: FutureWarning: pandas.util.testing is deprecated. Use the functions in the public API at pandas.testing instead.
  import pandas.util.testing as tm

from google.colab import drive #mount google drive
drive.mount('/content/drive')

Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3aietf%3awg%3aoauth%3a2.0%3aoob&response_type=code&scope=email%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdocs.test%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive.photos.readonly%20https%3a%2f%2fwww.googleapis.com%2fauth%2fpeopleapi.readonly

Enter your authorization code:
··········
Mounted at /content/drive

os.chdir('/content/drive/My Drive/Capstone My Data') #Change working directory

class_info_df = pd.read_csv('stage_2_detailed_class_info.csv') #Load csv files
train_labels_df = pd.read_csv('stage_2_train_labels.csv')

train_labels_df.shape #Print shape of train_labels_df

(30227, 6)

class_info_df.shape #Print shape of class_info_df

(30227, 2)

class_info_df.head(5) #print 5 records

train_labels_df.head(5) #print 5 records

1. Missing Values¶

def missing_data(data):
    total = data.isnull().sum().sort_values(ascending = False)
    percent = (data.isnull().sum()/data.isnull().count()*100).sort_values(ascending = False)
    return np.transpose(pd.concat([total, percent], axis=1, keys=['Total', 'Percent']))

missing_data(train_labels_df[train_labels_df['Target']==0]) #Normal

missing_data(train_labels_df[train_labels_df['Target']==1]) # Positive

missing_data(class_info_df)

2. Data distribution¶

f, ax = plt.subplots(1,1, figsize=(6,4))
total = float(len(class_info_df))
sns.countplot(class_info_df['class'],order = class_info_df['class'].value_counts().index, palette='Set3')
for p in ax.patches:
    height = p.get_height()
    ax.text(p.get_x()+p.get_width()/2.,
            height + 3,
            '{:1.2f}%'.format(100*height/total),
            ha="center") 
plt.show()

def get_distribution(data, feature):
    # Get the count for each label
    label_counts = data[feature].value_counts()
    # Get total number of samples
    total_samples = len(data)
    # Count the number of items in each class
    print("Feature: {}".format(feature))
    for i in range(len(label_counts)):
        label = label_counts.index[i]
        count = label_counts.values[i]
        percent = int((count / total_samples) * 10000) / 100
        print("{:<30s}:   {} or {}%".format(label, count, percent))

get_distribution(class_info_df, 'class')

Feature: class
No Lung Opacity / Not Normal  :   11821 or 39.1%
Lung Opacity                  :   9555 or 31.61%
Normal                        :   8851 or 29.28%

3. Merge two data frames¶

train_class_df = train_labels_df.merge(class_info_df, left_on='patientId', right_on='patientId', how='inner')

train_class_df.head(5)

4. Explore merged data set¶

fig, ax = plt.subplots(nrows=1,figsize=(12,6))
tmp = train_class_df.groupby('Target')['class'].value_counts()
df = pd.DataFrame(data={'Exams': tmp.values}, index=tmp.index).reset_index()
sns.barplot(ax=ax,x = 'Target', y='Exams',hue='class',data=df, palette='Set3')
plt.title("Class and Target")
plt.show()

Note -¶

Target = 1 is associated with class: Lung Opacity.

Target = 0 are either of class: Normal or class: No Lung Opacity / Not Normal.

5. Distribution of X, Y, height & width of Lung Opacity¶

target1 = train_class_df[train_class_df['Target']==1]
plt.figure()
fig, ax = plt.subplots(2,2,figsize=(12,12))
sns.distplot(target1['x'],kde=True,bins=50, color="red", ax=ax[0,0])
sns.distplot(target1['y'],kde=True,bins=50, color="blue", ax=ax[0,1])
sns.distplot(target1['width'],kde=True,bins=50, color="green", ax=ax[1,0])
sns.distplot(target1['height'],kde=True,bins=50, color="magenta", ax=ax[1,1])
locs, labels = plt.xticks()
plt.tick_params(axis='both', which='major', labelsize=12)
plt.show()

<Figure size 432x288 with 0 Axes>

6. Explore images¶

image_train_path = os.listdir('stage_2_train_images')
image_test_path = os.listdir('stage_2_test_images')
print("Number of images in train set:", len(image_train_path),"\nNumber of images in test set:", len(image_test_path))

Number of images in train set: 26684 
Number of images in test set: 3000

# Display image metadata
samplePatientID = list(train_class_df[:3].T.to_dict().values())[0]['patientId']
samplePatientID = samplePatientID+'.dcm'
dicom_file_path = os.path.join("stage_2_train_images/",samplePatientID)
dicom_file_dataset = dcm.read_file(dicom_file_path)
dicom_file_dataset

Dataset.file_meta -------------------------------
(0002, 0000) File Meta Information Group Length  UL: 202
(0002, 0001) File Meta Information Version       OB: b'\x00\x01'
(0002, 0002) Media Storage SOP Class UID         UI: Secondary Capture Image Storage
(0002, 0003) Media Storage SOP Instance UID      UI: 1.2.276.0.7230010.3.1.4.8323329.28530.1517874485.775526
(0002, 0010) Transfer Syntax UID                 UI: JPEG Baseline (Process 1)
(0002, 0012) Implementation Class UID            UI: 1.2.276.0.7230010.3.0.3.6.0
(0002, 0013) Implementation Version Name         SH: 'OFFIS_DCMTK_360'
-------------------------------------------------
(0008, 0005) Specific Character Set              CS: 'ISO_IR 100'
(0008, 0016) SOP Class UID                       UI: Secondary Capture Image Storage
(0008, 0018) SOP Instance UID                    UI: 1.2.276.0.7230010.3.1.4.8323329.28530.1517874485.775526
(0008, 0020) Study Date                          DA: '19010101'
(0008, 0030) Study Time                          TM: '000000.00'
(0008, 0050) Accession Number                    SH: ''
(0008, 0060) Modality                            CS: 'CR'
(0008, 0064) Conversion Type                     CS: 'WSD'
(0008, 0090) Referring Physician's Name          PN: ''
(0008, 103e) Series Description                  LO: 'view: PA'
(0010, 0010) Patient's Name                      PN: '0004cfab-14fd-4e49-80ba-63a80b6bddd6'
(0010, 0020) Patient ID                          LO: '0004cfab-14fd-4e49-80ba-63a80b6bddd6'
(0010, 0030) Patient's Birth Date                DA: ''
(0010, 0040) Patient's Sex                       CS: 'F'
(0010, 1010) Patient's Age                       AS: '51'
(0018, 0015) Body Part Examined                  CS: 'CHEST'
(0018, 5101) View Position                       CS: 'PA'
(0020, 000d) Study Instance UID                  UI: 1.2.276.0.7230010.3.1.2.8323329.28530.1517874485.775525
(0020, 000e) Series Instance UID                 UI: 1.2.276.0.7230010.3.1.3.8323329.28530.1517874485.775524
(0020, 0010) Study ID                            SH: ''
(0020, 0011) Series Number                       IS: "1"
(0020, 0013) Instance Number                     IS: "1"
(0020, 0020) Patient Orientation                 CS: ''
(0028, 0002) Samples per Pixel                   US: 1
(0028, 0004) Photometric Interpretation          CS: 'MONOCHROME2'
(0028, 0010) Rows                                US: 1024
(0028, 0011) Columns                             US: 1024
(0028, 0030) Pixel Spacing                       DS: [0.14300000000000002, 0.14300000000000002]
(0028, 0100) Bits Allocated                      US: 8
(0028, 0101) Bits Stored                         US: 8
(0028, 0102) High Bit                            US: 7
(0028, 0103) Pixel Representation                US: 0
(0028, 2110) Lossy Image Compression             CS: '01'
(0028, 2114) Lossy Image Compression Method      CS: 'ISO_10918_1'
(7fe0, 0010) Pixel Data                          OB: Array of 142006 elements

# Show dom images 
def show_dicom_images(data):
    img_data = list(data.T.to_dict().values())
    f, ax = plt.subplots(2,3, figsize=(16,12))
    for i,data_row in enumerate(img_data):
        patientImage = data_row['patientId']+'.dcm'
        imagePath = os.path.join("stage_2_train_images/",patientImage)
        data_row_img_data = dcm.read_file(imagePath)
        age = data_row_img_data.PatientAge
        sex = data_row_img_data.PatientSex
        #read dcm image
        data_row_img = dcm.dcmread(imagePath)
        ax[i//3, i%3].imshow(data_row_img.pixel_array, cmap=plt.cm.bone) 
        ax[i//3, i%3].axis('off')
        ax[i//3, i%3].set_title('ID: {}\nAge: {} Sex: {} Target: {}\nClass: {}\nWindow: {}:{}:{}:{}'.format(
                data_row['patientId'], age, sex, data_row['Target'], data_row['class'], 
                data_row['x'],data_row['y'],data_row['width'],data_row['height']))
    plt.show()

show_dicom_images(train_class_df[train_class_df['Target']==1].sample(6)) # for target 1

def show_dicom_images_with_boxes(data):
    img_data = list(data.T.to_dict().values())
    f, ax = plt.subplots(2,3, figsize=(16,12))
    for i,data_row in enumerate(img_data):
        patientImage = data_row['patientId']+'.dcm'
        imagePath = os.path.join("stage_2_train_images/",patientImage)
        data_row_img_data = dcm.read_file(imagePath)
        age = data_row_img_data.PatientAge
        sex = data_row_img_data.PatientSex
        data_row_img = dcm.dcmread(imagePath)
        ax[i//3, i%3].imshow(data_row_img.pixel_array, cmap=plt.cm.bone) 
        ax[i//3, i%3].axis('off')
        ax[i//3, i%3].set_title('ID: {}\nAge: {} Sex: {} Target: {}\nClass: {}'.format(
                data_row['patientId'], age, sex, data_row['Target'], data_row['class']))
        rows = train_class_df[train_class_df['patientId']==data_row['patientId']]
        box_data = list(rows.T.to_dict().values())
        for j, row in enumerate(box_data):
            ax[i//3, i%3].add_patch(Rectangle(xy=(row['x'], row['y']),
                        width=row['width'],height=row['height'], 
                        color="red",alpha = 0.1))   
    plt.show()

show_dicom_images_with_boxes(train_class_df[train_class_df['Target']==1].sample(6))

show_dicom_images(train_class_df[train_class_df['Target']==0].sample(6)) # Display images with target 0

EDA Summary¶

We have total 26,684 images in training data set & 3000 images in testing dataset.
Target = 1 is associated with class: Lung Opacity.
Target = 0 are either of class: Normal or class: No Lung Opacity
There is no x,y,width & height value for records with Target 0.
There are two gaussian for x point distrubution graph. This respresent Lung opacity are in both left & righ lungs in given dataset.

7. Load pneumonia locations¶

Table contains [filename : pneumonia location] pairs per row.

If a filename contains multiple pneumonia, the table contains multiple rows with the same filename but different pneumonia locations.
If a filename contains no pneumonia it contains a single row with an empty pneumonia location.

The code below loads the table and transforms it into a dictionary.

The dictionary uses the filename as key and a list of pneumonia locations in that filename as value.
If a filename is not present in the dictionary it means that it contains no pneumonia.

# empty dictionary
pneumonia_locations = {}
# load table
with open(os.path.join('stage_2_train_labels.csv'), mode='r') as infile:
    # open reader
    reader = csv.reader(infile)
    # skip header
    next(reader, None)
    # loop through rows
    for rows in reader:
        # retrieve information
        filename = rows[0]
        location = rows[1:5]
        pneumonia = rows[5]
        # if row contains pneumonia add label to dictionary
        # which contains a list of pneumonia locations per filename
        if pneumonia == '1':
            # convert string to float to int
            location = [int(float(i)) for i in location]
            # save pneumonia location in dictionary
            if filename in pneumonia_locations:
                pneumonia_locations[filename].append(location)
            else:
                pneumonia_locations[filename] = [location]

train_class_df["filenames"]=train_class_df['patientId']+'.dcm' # Create one more column for file name

train_class_df.head()

POSITIVE = train_class_df[train_class_df['Target']==1]
NEGATIVE = train_class_df[train_class_df['Target']==0]

POSITIVE.shape

(16957, 8)

NEGATIVE.shape

(20672, 8)

#Use only 1.5K records per class for training.. as its taking time for training
records_for_training=1500
POSITIVE = POSITIVE[:records_for_training]
NEGATIVE = NEGATIVE[:records_for_training]

training_df=pd.concat([POSITIVE,NEGATIVE])

training_df.head()

filenames = list(training_df["filenames"])

random.shuffle(filenames)

# split into train and validation small set
n_valid_samples = 600
total_records_for_training=3000
train_filenames = filenames[n_valid_samples:total_records_for_training]
valid_filenames = filenames[:n_valid_samples]

len(valid_filenames)

600

len(train_filenames)

2400

Data generator¶

The dataset is too large to fit into memory, so we need to create a generator that loads data on the fly.

The generator takes in some filenames, batch_size and other parameters.
The generator outputs a random batch of numpy images and numpy masks.

class generator(keras.utils.Sequence):
    
    def __init__(self, folder, filenames, pneumonia_locations=None, batch_size=2, image_size=128, shuffle=True, augment=False, predict=False):
        self.folder = folder
        self.filenames = filenames
        self.pneumonia_locations = pneumonia_locations
        self.batch_size = batch_size
        self.image_size = image_size
        self.shuffle = shuffle
        self.augment = augment
        self.predict = predict
        self.on_epoch_end()
        
    def __load__(self, filename):
        # load dicom file as numpy array
        img = pydicom.dcmread(os.path.join(self.folder, filename)).pixel_array
        # create empty mask
        msk = np.zeros(img.shape)
        # get filename without extension
        filename = filename.split('.')[0]
        # if image contains pneumonia
        if filename in self.pneumonia_locations:
            # loop through pneumonia
            for location in self.pneumonia_locations[filename]:
                # add 1's at the location of the pneumonia
                x, y, w, h = location
                msk[y:y+h, x:x+w] = 1
        # resize both image and mask
        img = resize(img, (self.image_size, self.image_size), mode='reflect')
        msk = resize(msk, (self.image_size, self.image_size), mode='reflect') > 0.5
        # if augment then horizontal flip half the time
        if self.augment and random.random() > 0.5:
            img = np.fliplr(img)
            msk = np.fliplr(msk)
        # add trailing channel dimension
        img = np.expand_dims(img, -1)
        msk = np.expand_dims(msk, -1)
        return img, msk
    
    def __loadpredict__(self, filename):
        # load dicom file as numpy array
        img = pydicom.dcmread(os.path.join(self.folder, filename)).pixel_array
        # resize image
        img = resize(img, (self.image_size, self.image_size), mode='reflect')
        # add trailing channel dimension
        img = np.expand_dims(img, -1)
        return img
        
    def __getitem__(self, index):
        # select batch
        filenames = self.filenames[index*self.batch_size:(index+1)*self.batch_size]
        # predict mode: return images and filenames
        if self.predict:
            # load files
            imgs = [self.__loadpredict__(filename) for filename in filenames]
            # create numpy batch
            imgs = np.array(imgs)
            return imgs, filenames
        # train mode: return images and masks
        else:
            # load files
            items = [self.__load__(filename) for filename in filenames]
            # unzip images and masks
            imgs, msks = zip(*items)
            # create numpy batch
            imgs = np.array(imgs)
            msks = np.array(msks)
            return imgs, msks
        
    def on_epoch_end(self):
        if self.shuffle:
            random.shuffle(self.filenames)
        
    def __len__(self):
        if self.predict:
            # return everything
            return int(np.ceil(len(self.filenames) / self.batch_size))
        else:
            # return full batches only
            return int(len(self.filenames) / self.batch_size)

ResNet¶

In traditional neural networks, more layers mean a better network but because of the vanishing gradient problem, weights of the first layer won’t be updated correctly through the back-propagation. As the error gradient is back-propagated to earlier layers, repeated multiplication makes the gradient small. Thus, with more layers in the networks, its performance gets saturated and starts decreasing rapidly. Res-Net solves this problem by using the identity matrix. When the back-propagation is done through identity function, the gradient will be multiplied only by 1. This preserves the input and avoids any loss in the information.¶

def create_downsample(channels, inputs):
    x = keras.layers.BatchNormalization(momentum=0.9)(inputs)
    x = keras.layers.LeakyReLU(0)(x)
    x = keras.layers.Conv2D(channels, 1, padding='same', use_bias=False)(x)
    x = keras.layers.MaxPool2D(2)(x)
    return x

def create_resblock(channels, inputs):
    x = keras.layers.BatchNormalization(momentum=0.9)(inputs)
    x = keras.layers.LeakyReLU(0)(x)
    x = keras.layers.Conv2D(channels, 3, padding='same', use_bias=False)(x)
    x = keras.layers.BatchNormalization(momentum=0.9)(x)
    x = keras.layers.LeakyReLU(0)(x)
    x = keras.layers.Conv2D(channels, 3, padding='same', use_bias=False)(x)
    return keras.layers.add([x, inputs])

def create_network(input_size, channels, n_blocks=2, depth=4):
    # input
    inputs = keras.Input(shape=(input_size, input_size, 1))
    x = keras.layers.Conv2D(channels, 3, padding='same', use_bias=False)(inputs)
    # residual blocks
    for d in range(depth):
        channels = channels * 2
        x = create_downsample(channels, x)
        for b in range(n_blocks):
            x = create_resblock(channels, x)
    # output
    x = keras.layers.BatchNormalization(momentum=0.9)(x)
    x = keras.layers.LeakyReLU(0)(x)
    x = keras.layers.Conv2D(1, 1, activation='sigmoid')(x)
    outputs = keras.layers.UpSampling2D(2**depth)(x)
    model = keras.Model(inputs=inputs, outputs=outputs)
    return model

Train Network¶

# define iou or jaccard loss function
def iou_loss(y_true, y_pred):
    y_true = tf.reshape(y_true, [-1])
    y_pred = tf.reshape(y_pred, [-1])
    intersection = tf.reduce_sum(y_true * y_pred)
    score = (intersection + 1.) / (tf.reduce_sum(y_true) + tf.reduce_sum(y_pred) - intersection + 1.)
    return 1 - score

# combine bce loss and iou loss
def iou_bce_loss(y_true, y_pred):
    return 0.5 * keras.losses.binary_crossentropy(y_true, y_pred) + 0.5 * iou_loss(y_true, y_pred)

# mean iou as a metric
def mean_iou(y_true, y_pred):
    y_pred = tf.round(y_pred)
    intersect = tf.reduce_sum(y_true * y_pred, axis=[1, 2, 3])
    union = tf.reduce_sum(y_true, axis=[1, 2, 3]) + tf.reduce_sum(y_pred, axis=[1, 2, 3])
    smooth = tf.ones(tf.shape(intersect))
    return tf.reduce_mean((intersect + smooth) / (union - intersect + smooth))

# cosine learning rate annealing
def cosine_annealing(x):
    lr = 0.001
    epochs = 25
    return lr*(np.cos(np.pi*x/epochs)+1.)/2
learning_rate = tf.keras.callbacks.LearningRateScheduler(cosine_annealing)

# create network and compiler
model = create_network(input_size=256, channels=32, n_blocks=2, depth=4)
model.compile(optimizer='adam',
              loss=iou_bce_loss,
              metrics=['accuracy', mean_iou])
# create train and validation generators
folder = 'stage_2_train_images'
train_gen = generator(folder, train_filenames, pneumonia_locations, batch_size=32, image_size=256, shuffle=True, augment=True, predict=False)
valid_gen = generator(folder, valid_filenames, pneumonia_locations, batch_size=32, image_size=256, shuffle=False, predict=False)

from tensorflow.keras.callbacks import ModelCheckpoint
checkpoint_file="4000_checkpoint.hdf5"
checkpoint = ModelCheckpoint(checkpoint_file,monitor='accuracy',verbose=1,save_best_only=True,mode='max',save_freq=1)

history = model.fit_generator(train_gen, validation_data=valid_gen, callbacks=[learning_rate,checkpoint], epochs=6, workers=8, use_multiprocessing=True)

Epoch 1/6
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 1/78 [..............................] - ETA: 0s - loss: 0.8666 - accuracy: 0.4816 - mean_iou: 0.0651WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 2/78 [..............................] - ETA: 27:05 - loss: 1.0757 - accuracy: 0.4222 - mean_iou: 0.0997WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 3/78 [>.............................] - ETA: 35:40 - loss: 0.9744 - accuracy: 0.4635 - mean_iou: 0.0997WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 4/78 [>.............................] - ETA: 39:31 - loss: 0.9097 - accuracy: 0.5661 - mean_iou: 0.0868WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 5/78 [>.............................] - ETA: 41:35 - loss: 0.8582 - accuracy: 0.6218 - mean_iou: 0.1023WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 6/78 [=>............................] - ETA: 42:43 - loss: 0.8207 - accuracy: 0.6519 - mean_iou: 0.1134WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 7/78 [=>............................] - ETA: 43:17 - loss: 0.7845 - accuracy: 0.6808 - mean_iou: 0.1382WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 8/78 [==>...........................] - ETA: 43:34 - loss: 0.7561 - accuracy: 0.7061 - mean_iou: 0.1533WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 9/78 [==>...........................] - ETA: 43:39 - loss: 0.7360 - accuracy: 0.7238 - mean_iou: 0.1619WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
10/78 [==>...........................] - ETA: 43:39 - loss: 0.7179 - accuracy: 0.7417 - mean_iou: 0.1839WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
11/78 [===>..........................] - ETA: 43:29 - loss: 0.7037 - accuracy: 0.7559 - mean_iou: 0.2011WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
12/78 [===>..........................] - ETA: 43:19 - loss: 0.6877 - accuracy: 0.7684 - mean_iou: 0.2167WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
13/78 [====>.........................] - ETA: 42:59 - loss: 0.6760 - accuracy: 0.7775 - mean_iou: 0.2323WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
14/78 [====>.........................] - ETA: 42:37 - loss: 0.6653 - accuracy: 0.7856 - mean_iou: 0.2480WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
15/78 [====>.........................] - ETA: 42:13 - loss: 0.6544 - accuracy: 0.7943 - mean_iou: 0.2657WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
16/78 [=====>........................] - ETA: 41:45 - loss: 0.6449 - accuracy: 0.8007 - mean_iou: 0.2796WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
17/78 [=====>........................] - ETA: 41:16 - loss: 0.6402 - accuracy: 0.8047 - mean_iou: 0.2797WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
18/78 [=====>........................] - ETA: 40:45 - loss: 0.6329 - accuracy: 0.8084 - mean_iou: 0.2902WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
19/78 [======>.......................] - ETA: 40:11 - loss: 0.6283 - accuracy: 0.8107 - mean_iou: 0.2988WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
20/78 [======>.......................] - ETA: 39:37 - loss: 0.6236 - accuracy: 0.8132 - mean_iou: 0.3008WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
21/78 [=======>......................] - ETA: 39:02 - loss: 0.6196 - accuracy: 0.8156 - mean_iou: 0.2990WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
22/78 [=======>......................] - ETA: 38:26 - loss: 0.6152 - accuracy: 0.8190 - mean_iou: 0.3011WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
23/78 [=======>......................] - ETA: 37:49 - loss: 0.6107 - accuracy: 0.8222 - mean_iou: 0.3071WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
24/78 [========>.....................] - ETA: 37:12 - loss: 0.6066 - accuracy: 0.8245 - mean_iou: 0.3117WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
25/78 [========>.....................] - ETA: 36:34 - loss: 0.6027 - accuracy: 0.8275 - mean_iou: 0.3165WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
26/78 [=========>....................] - ETA: 35:55 - loss: 0.5992 - accuracy: 0.8297 - mean_iou: 0.3215WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
27/78 [=========>....................] - ETA: 35:17 - loss: 0.5960 - accuracy: 0.8321 - mean_iou: 0.3246WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
28/78 [=========>....................] - ETA: 34:38 - loss: 0.5926 - accuracy: 0.8342 - mean_iou: 0.3319WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
29/78 [==========>...................] - ETA: 34:01 - loss: 0.5922 - accuracy: 0.8358 - mean_iou: 0.3343WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
30/78 [==========>...................] - ETA: 33:22 - loss: 0.5893 - accuracy: 0.8385 - mean_iou: 0.3381WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
31/78 [==========>...................] - ETA: 32:42 - loss: 0.5878 - accuracy: 0.8403 - mean_iou: 0.3406WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
32/78 [===========>..................] - ETA: 32:02 - loss: 0.5844 - accuracy: 0.8427 - mean_iou: 0.3476WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
33/78 [===========>..................] - ETA: 31:22 - loss: 0.5812 - accuracy: 0.8449 - mean_iou: 0.3542WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
34/78 [============>.................] - ETA: 30:42 - loss: 0.5774 - accuracy: 0.8471 - mean_iou: 0.3605WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
35/78 [============>.................] - ETA: 30:02 - loss: 0.5738 - accuracy: 0.8492 - mean_iou: 0.3645WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
36/78 [============>.................] - ETA: 29:22 - loss: 0.5731 - accuracy: 0.8501 - mean_iou: 0.3668WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
37/78 [=============>................] - ETA: 28:41 - loss: 0.5707 - accuracy: 0.8518 - mean_iou: 0.3681WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
38/78 [=============>................] - ETA: 28:01 - loss: 0.5696 - accuracy: 0.8529 - mean_iou: 0.3701WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
39/78 [==============>...............] - ETA: 27:19 - loss: 0.5681 - accuracy: 0.8541 - mean_iou: 0.3699WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
40/78 [==============>...............] - ETA: 26:38 - loss: 0.5656 - accuracy: 0.8554 - mean_iou: 0.3736WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
41/78 [==============>...............] - ETA: 25:57 - loss: 0.5636 - accuracy: 0.8566 - mean_iou: 0.3739WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
42/78 [===============>..............] - ETA: 25:16 - loss: 0.5611 - accuracy: 0.8575 - mean_iou: 0.3771WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
43/78 [===============>..............] - ETA: 24:36 - loss: 0.5585 - accuracy: 0.8586 - mean_iou: 0.3822WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
44/78 [===============>..............] - ETA: 23:54 - loss: 0.5558 - accuracy: 0.8597 - mean_iou: 0.3858WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
45/78 [================>.............] - ETA: 23:13 - loss: 0.5546 - accuracy: 0.8603 - mean_iou: 0.3862WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
46/78 [================>.............] - ETA: 22:32 - loss: 0.5536 - accuracy: 0.8613 - mean_iou: 0.3890WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
47/78 [=================>............] - ETA: 21:50 - loss: 0.5525 - accuracy: 0.8624 - mean_iou: 0.3897WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
48/78 [=================>............] - ETA: 21:09 - loss: 0.5521 - accuracy: 0.8634 - mean_iou: 0.3918WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
49/78 [=================>............] - ETA: 20:27 - loss: 0.5516 - accuracy: 0.8643 - mean_iou: 0.3927WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
50/78 [==================>...........] - ETA: 19:46 - loss: 0.5510 - accuracy: 0.8655 - mean_iou: 0.3933WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
51/78 [==================>...........] - ETA: 19:04 - loss: 0.5492 - accuracy: 0.8667 - mean_iou: 0.3985WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
52/78 [===================>..........] - ETA: 18:22 - loss: 0.5489 - accuracy: 0.8672 - mean_iou: 0.3987WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
53/78 [===================>..........] - ETA: 17:40 - loss: 0.5495 - accuracy: 0.8676 - mean_iou: 0.3987WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
54/78 [===================>..........] - ETA: 16:58 - loss: 0.5480 - accuracy: 0.8684 - mean_iou: 0.3998WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
55/78 [====================>.........] - ETA: 16:16 - loss: 0.5470 - accuracy: 0.8691 - mean_iou: 0.3999WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
56/78 [====================>.........] - ETA: 15:34 - loss: 0.5453 - accuracy: 0.8700 - mean_iou: 0.4010WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
57/78 [====================>.........] - ETA: 14:52 - loss: 0.5440 - accuracy: 0.8709 - mean_iou: 0.4038WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
58/78 [=====================>........] - ETA: 14:10 - loss: 0.5432 - accuracy: 0.8715 - mean_iou: 0.4032WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
59/78 [=====================>........] - ETA: 13:28 - loss: 0.5423 - accuracy: 0.8723 - mean_iou: 0.4035WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
60/78 [======================>.......] - ETA: 12:45 - loss: 0.5416 - accuracy: 0.8727 - mean_iou: 0.4043WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
61/78 [======================>.......] - ETA: 12:03 - loss: 0.5406 - accuracy: 0.8735 - mean_iou: 0.4047WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
62/78 [======================>.......] - ETA: 11:21 - loss: 0.5407 - accuracy: 0.8735 - mean_iou: 0.4044WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
63/78 [=======================>......] - ETA: 10:38 - loss: 0.5406 - accuracy: 0.8737 - mean_iou: 0.4042WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
64/78 [=======================>......] - ETA: 9:56 - loss: 0.5398 - accuracy: 0.8741 - mean_iou: 0.4044 WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
65/78 [========================>.....] - ETA: 9:13 - loss: 0.5390 - accuracy: 0.8745 - mean_iou: 0.4051WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
66/78 [========================>.....] - ETA: 8:31 - loss: 0.5386 - accuracy: 0.8749 - mean_iou: 0.4057WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
67/78 [========================>.....] - ETA: 7:48 - loss: 0.5378 - accuracy: 0.8754 - mean_iou: 0.4055WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
68/78 [=========================>....] - ETA: 7:05 - loss: 0.5371 - accuracy: 0.8759 - mean_iou: 0.4064WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
69/78 [=========================>....] - ETA: 6:23 - loss: 0.5377 - accuracy: 0.8760 - mean_iou: 0.4055WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
70/78 [=========================>....] - ETA: 5:40 - loss: 0.5366 - accuracy: 0.8766 - mean_iou: 0.4066WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
71/78 [==========================>...] - ETA: 4:57 - loss: 0.5358 - accuracy: 0.8768 - mean_iou: 0.4068WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
72/78 [==========================>...] - ETA: 4:14 - loss: 0.5352 - accuracy: 0.8771 - mean_iou: 0.4066WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
73/78 [===========================>..] - ETA: 3:32 - loss: 0.5348 - accuracy: 0.8771 - mean_iou: 0.4065WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
74/78 [===========================>..] - ETA: 2:49 - loss: 0.5347 - accuracy: 0.8773 - mean_iou: 0.4083WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
75/78 [===========================>..] - ETA: 2:07 - loss: 0.5346 - accuracy: 0.8774 - mean_iou: 0.4088WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
76/78 [============================>.] - ETA: 1:24 - loss: 0.5334 - accuracy: 0.8779 - mean_iou: 0.4105WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
77/78 [============================>.] - ETA: 42s - loss: 0.5324 - accuracy: 0.8783 - mean_iou: 0.4120 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
78/78 [==============================] - ETA: 0s - loss: 0.5318 - accuracy: 0.8788 - mean_iou: 0.4122 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
78/78 [==============================] - 3584s 46s/step - loss: 0.5318 - accuracy: 0.8788 - mean_iou: 0.4122 - val_loss: 0.4704 - val_accuracy: 0.9058 - val_mean_iou: 0.4706 - lr: 0.0010
Epoch 2/6
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 1/78 [..............................] - ETA: 0s - loss: 0.4139 - accuracy: 0.9449 - mean_iou: 0.5732WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 2/78 [..............................] - ETA: 27:32 - loss: 0.4606 - accuracy: 0.9219 - mean_iou: 0.4910WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 3/78 [>.............................] - ETA: 36:35 - loss: 0.4642 - accuracy: 0.9183 - mean_iou: 0.4922WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 4/78 [>.............................] - ETA: 40:36 - loss: 0.4729 - accuracy: 0.9154 - mean_iou: 0.4893WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 5/78 [>.............................] - ETA: 42:37 - loss: 0.4683 - accuracy: 0.9126 - mean_iou: 0.4963WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 6/78 [=>............................] - ETA: 43:42 - loss: 0.4743 - accuracy: 0.9103 - mean_iou: 0.4940WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 7/78 [=>............................] - ETA: 44:17 - loss: 0.4828 - accuracy: 0.9067 - mean_iou: 0.4910WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 8/78 [==>...........................] - ETA: 44:33 - loss: 0.4852 - accuracy: 0.9060 - mean_iou: 0.4775WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 9/78 [==>...........................] - ETA: 44:33 - loss: 0.4809 - accuracy: 0.9082 - mean_iou: 0.4862WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
10/78 [==>...........................] - ETA: 44:29 - loss: 0.4839 - accuracy: 0.9062 - mean_iou: 0.4833WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
11/78 [===>..........................] - ETA: 44:18 - loss: 0.4800 - accuracy: 0.9085 - mean_iou: 0.4912WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
12/78 [===>..........................] - ETA: 44:03 - loss: 0.4778 - accuracy: 0.9102 - mean_iou: 0.4925WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
13/78 [====>.........................] - ETA: 43:41 - loss: 0.4801 - accuracy: 0.9098 - mean_iou: 0.4882WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
14/78 [====>.........................] - ETA: 43:15 - loss: 0.4840 - accuracy: 0.9108 - mean_iou: 0.4840WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
15/78 [====>.........................] - ETA: 42:46 - loss: 0.4825 - accuracy: 0.9118 - mean_iou: 0.4878WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
16/78 [=====>........................] - ETA: 42:17 - loss: 0.4786 - accuracy: 0.9137 - mean_iou: 0.4965WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
17/78 [=====>........................] - ETA: 41:44 - loss: 0.4778 - accuracy: 0.9139 - mean_iou: 0.4927WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
18/78 [=====>........................] - ETA: 41:11 - loss: 0.4754 - accuracy: 0.9146 - mean_iou: 0.4973WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
19/78 [======>.......................] - ETA: 40:36 - loss: 0.4759 - accuracy: 0.9137 - mean_iou: 0.4958WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
20/78 [======>.......................] - ETA: 40:00 - loss: 0.4754 - accuracy: 0.9143 - mean_iou: 0.4943WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
21/78 [=======>......................] - ETA: 39:25 - loss: 0.4779 - accuracy: 0.9133 - mean_iou: 0.4924WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
22/78 [=======>......................] - ETA: 38:47 - loss: 0.4759 - accuracy: 0.9135 - mean_iou: 0.4977WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
23/78 [=======>......................] - ETA: 38:09 - loss: 0.4756 - accuracy: 0.9129 - mean_iou: 0.4951WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
24/78 [========>.....................] - ETA: 37:31 - loss: 0.4771 - accuracy: 0.9125 - mean_iou: 0.4904WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
25/78 [========>.....................] - ETA: 36:55 - loss: 0.4758 - accuracy: 0.9129 - mean_iou: 0.4848WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
26/78 [=========>....................] - ETA: 36:31 - loss: 0.4751 - accuracy: 0.9130 - mean_iou: 0.4848WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
27/78 [=========>....................] - ETA: 36:05 - loss: 0.4778 - accuracy: 0.9122 - mean_iou: 0.4841WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
28/78 [=========>....................] - ETA: 35:40 - loss: 0.4776 - accuracy: 0.9121 - mean_iou: 0.4841WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
29/78 [==========>...................] - ETA: 35:10 - loss: 0.4790 - accuracy: 0.9111 - mean_iou: 0.4817WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
30/78 [==========>...................] - ETA: 34:29 - loss: 0.4765 - accuracy: 0.9119 - mean_iou: 0.4830WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
31/78 [==========>...................] - ETA: 33:50 - loss: 0.4747 - accuracy: 0.9122 - mean_iou: 0.4854WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
32/78 [===========>..................] - ETA: 33:20 - loss: 0.4729 - accuracy: 0.9123 - mean_iou: 0.4882WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
33/78 [===========>..................] - ETA: 32:40 - loss: 0.4726 - accuracy: 0.9120 - mean_iou: 0.4878WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
34/78 [============>.................] - ETA: 31:58 - loss: 0.4733 - accuracy: 0.9116 - mean_iou: 0.4857WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
35/78 [============>.................] - ETA: 31:15 - loss: 0.4742 - accuracy: 0.9113 - mean_iou: 0.4842WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
36/78 [============>.................] - ETA: 30:33 - loss: 0.4728 - accuracy: 0.9117 - mean_iou: 0.4862WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
37/78 [=============>................] - ETA: 29:50 - loss: 0.4730 - accuracy: 0.9118 - mean_iou: 0.4857WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
38/78 [=============>................] - ETA: 29:06 - loss: 0.4739 - accuracy: 0.9119 - mean_iou: 0.4845WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
39/78 [==============>...............] - ETA: 28:22 - loss: 0.4739 - accuracy: 0.9119 - mean_iou: 0.4867WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
40/78 [==============>...............] - ETA: 27:39 - loss: 0.4722 - accuracy: 0.9128 - mean_iou: 0.4870WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
41/78 [==============>...............] - ETA: 26:55 - loss: 0.4721 - accuracy: 0.9133 - mean_iou: 0.4855WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
42/78 [===============>..............] - ETA: 26:11 - loss: 0.4706 - accuracy: 0.9141 - mean_iou: 0.4878WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
43/78 [===============>..............] - ETA: 25:28 - loss: 0.4710 - accuracy: 0.9142 - mean_iou: 0.4876WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
44/78 [===============>..............] - ETA: 24:44 - loss: 0.4720 - accuracy: 0.9139 - mean_iou: 0.4836WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
45/78 [================>.............] - ETA: 24:00 - loss: 0.4717 - accuracy: 0.9140 - mean_iou: 0.4840WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
46/78 [================>.............] - ETA: 23:16 - loss: 0.4699 - accuracy: 0.9144 - mean_iou: 0.4870WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
47/78 [=================>............] - ETA: 22:32 - loss: 0.4697 - accuracy: 0.9143 - mean_iou: 0.4881WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
48/78 [=================>............] - ETA: 21:49 - loss: 0.4716 - accuracy: 0.9136 - mean_iou: 0.4872WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
49/78 [=================>............] - ETA: 21:05 - loss: 0.4729 - accuracy: 0.9133 - mean_iou: 0.4860WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
50/78 [==================>...........] - ETA: 20:21 - loss: 0.4729 - accuracy: 0.9132 - mean_iou: 0.4848WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
51/78 [==================>...........] - ETA: 19:37 - loss: 0.4727 - accuracy: 0.9131 - mean_iou: 0.4816WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
52/78 [===================>..........] - ETA: 18:53 - loss: 0.4722 - accuracy: 0.9131 - mean_iou: 0.4805WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
53/78 [===================>..........] - ETA: 18:10 - loss: 0.4723 - accuracy: 0.9130 - mean_iou: 0.4785WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
54/78 [===================>..........] - ETA: 17:26 - loss: 0.4723 - accuracy: 0.9131 - mean_iou: 0.4772WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
55/78 [====================>.........] - ETA: 16:42 - loss: 0.4720 - accuracy: 0.9133 - mean_iou: 0.4783WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
56/78 [====================>.........] - ETA: 15:59 - loss: 0.4707 - accuracy: 0.9139 - mean_iou: 0.4818WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
57/78 [====================>.........] - ETA: 15:15 - loss: 0.4706 - accuracy: 0.9140 - mean_iou: 0.4828WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
58/78 [=====================>........] - ETA: 14:31 - loss: 0.4713 - accuracy: 0.9135 - mean_iou: 0.4829WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
59/78 [=====================>........] - ETA: 13:48 - loss: 0.4708 - accuracy: 0.9134 - mean_iou: 0.4828WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
60/78 [======================>.......] - ETA: 13:05 - loss: 0.4698 - accuracy: 0.9134 - mean_iou: 0.4847WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
61/78 [======================>.......] - ETA: 12:21 - loss: 0.4693 - accuracy: 0.9133 - mean_iou: 0.4850WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
62/78 [======================>.......] - ETA: 11:37 - loss: 0.4689 - accuracy: 0.9130 - mean_iou: 0.4850WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
63/78 [=======================>......] - ETA: 10:53 - loss: 0.4692 - accuracy: 0.9127 - mean_iou: 0.4828WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
64/78 [=======================>......] - ETA: 10:10 - loss: 0.4696 - accuracy: 0.9123 - mean_iou: 0.4821WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
65/78 [========================>.....] - ETA: 9:26 - loss: 0.4706 - accuracy: 0.9119 - mean_iou: 0.4806 WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
66/78 [========================>.....] - ETA: 8:42 - loss: 0.4706 - accuracy: 0.9117 - mean_iou: 0.4790WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
67/78 [========================>.....] - ETA: 7:59 - loss: 0.4703 - accuracy: 0.9117 - mean_iou: 0.4788WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
68/78 [=========================>....] - ETA: 7:15 - loss: 0.4696 - accuracy: 0.9119 - mean_iou: 0.4804WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
69/78 [=========================>....] - ETA: 6:31 - loss: 0.4701 - accuracy: 0.9120 - mean_iou: 0.4799WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
70/78 [=========================>....] - ETA: 5:47 - loss: 0.4694 - accuracy: 0.9123 - mean_iou: 0.4820WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
71/78 [==========================>...] - ETA: 5:03 - loss: 0.4696 - accuracy: 0.9124 - mean_iou: 0.4811WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
72/78 [==========================>...] - ETA: 4:20 - loss: 0.4696 - accuracy: 0.9124 - mean_iou: 0.4812WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
73/78 [===========================>..] - ETA: 3:36 - loss: 0.4700 - accuracy: 0.9123 - mean_iou: 0.4813WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
74/78 [===========================>..] - ETA: 2:53 - loss: 0.4698 - accuracy: 0.9123 - mean_iou: 0.4810WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
75/78 [===========================>..] - ETA: 2:09 - loss: 0.4695 - accuracy: 0.9124 - mean_iou: 0.4814WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
76/78 [============================>.] - ETA: 1:26 - loss: 0.4691 - accuracy: 0.9124 - mean_iou: 0.4814WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
77/78 [============================>.] - ETA: 43s - loss: 0.4690 - accuracy: 0.9123 - mean_iou: 0.4813 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
78/78 [==============================] - ETA: 0s - loss: 0.4690 - accuracy: 0.9121 - mean_iou: 0.4808 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
78/78 [==============================] - 3646s 47s/step - loss: 0.4690 - accuracy: 0.9121 - mean_iou: 0.4808 - val_loss: 0.5787 - val_accuracy: 0.8078 - val_mean_iou: 0.2066 - lr: 9.9606e-04
Epoch 3/6
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 1/78 [..............................] - ETA: 0s - loss: 0.4308 - accuracy: 0.9047 - mean_iou: 0.5029WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 2/78 [..............................] - ETA: 27:29 - loss: 0.4367 - accuracy: 0.9123 - mean_iou: 0.5125WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 3/78 [>.............................] - ETA: 36:11 - loss: 0.4345 - accuracy: 0.9120 - mean_iou: 0.4759WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 4/78 [>.............................] - ETA: 40:09 - loss: 0.4439 - accuracy: 0.9114 - mean_iou: 0.4574WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 5/78 [>.............................] - ETA: 42:14 - loss: 0.4524 - accuracy: 0.9102 - mean_iou: 0.4342WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 6/78 [=>............................] - ETA: 43:39 - loss: 0.4577 - accuracy: 0.9075 - mean_iou: 0.4260WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 7/78 [=>............................] - ETA: 44:18 - loss: 0.4579 - accuracy: 0.9087 - mean_iou: 0.4449WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 8/78 [==>...........................] - ETA: 44:36 - loss: 0.4619 - accuracy: 0.9086 - mean_iou: 0.4421WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 9/78 [==>...........................] - ETA: 44:38 - loss: 0.4573 - accuracy: 0.9099 - mean_iou: 0.4502WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
10/78 [==>...........................] - ETA: 44:33 - loss: 0.4611 - accuracy: 0.9098 - mean_iou: 0.4535WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
11/78 [===>..........................] - ETA: 44:19 - loss: 0.4624 - accuracy: 0.9102 - mean_iou: 0.4489WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
12/78 [===>..........................] - ETA: 44:04 - loss: 0.4587 - accuracy: 0.9112 - mean_iou: 0.4523WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
13/78 [====>.........................] - ETA: 43:41 - loss: 0.4563 - accuracy: 0.9124 - mean_iou: 0.4660WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
14/78 [====>.........................] - ETA: 43:17 - loss: 0.4559 - accuracy: 0.9132 - mean_iou: 0.4685WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
15/78 [====>.........................] - ETA: 42:51 - loss: 0.4551 - accuracy: 0.9143 - mean_iou: 0.4713WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
16/78 [=====>........................] - ETA: 42:22 - loss: 0.4563 - accuracy: 0.9144 - mean_iou: 0.4697WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
17/78 [=====>........................] - ETA: 41:50 - loss: 0.4540 - accuracy: 0.9160 - mean_iou: 0.4727WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
18/78 [=====>........................] - ETA: 41:19 - loss: 0.4543 - accuracy: 0.9162 - mean_iou: 0.4756WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
19/78 [======>.......................] - ETA: 40:45 - loss: 0.4517 - accuracy: 0.9170 - mean_iou: 0.4844WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
20/78 [======>.......................] - ETA: 40:09 - loss: 0.4536 - accuracy: 0.9168 - mean_iou: 0.4852WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
21/78 [=======>......................] - ETA: 39:33 - loss: 0.4516 - accuracy: 0.9174 - mean_iou: 0.4923WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
22/78 [=======>......................] - ETA: 38:56 - loss: 0.4478 - accuracy: 0.9185 - mean_iou: 0.5001WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
23/78 [=======>......................] - ETA: 38:19 - loss: 0.4460 - accuracy: 0.9187 - mean_iou: 0.5065WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
24/78 [========>.....................] - ETA: 37:51 - loss: 0.4448 - accuracy: 0.9192 - mean_iou: 0.5001WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
25/78 [========>.....................] - ETA: 37:13 - loss: 0.4437 - accuracy: 0.9192 - mean_iou: 0.5019WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
26/78 [=========>....................] - ETA: 36:35 - loss: 0.4419 - accuracy: 0.9195 - mean_iou: 0.5061WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
27/78 [=========>....................] - ETA: 35:55 - loss: 0.4442 - accuracy: 0.9185 - mean_iou: 0.5043WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
28/78 [=========>....................] - ETA: 35:16 - loss: 0.4426 - accuracy: 0.9190 - mean_iou: 0.5077WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
29/78 [==========>...................] - ETA: 34:36 - loss: 0.4422 - accuracy: 0.9190 - mean_iou: 0.5110WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
30/78 [==========>...................] - ETA: 33:56 - loss: 0.4416 - accuracy: 0.9191 - mean_iou: 0.5101WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
31/78 [==========>...................] - ETA: 33:15 - loss: 0.4405 - accuracy: 0.9191 - mean_iou: 0.5140WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
32/78 [===========>..................] - ETA: 32:35 - loss: 0.4403 - accuracy: 0.9192 - mean_iou: 0.5141WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
33/78 [===========>..................] - ETA: 31:54 - loss: 0.4423 - accuracy: 0.9182 - mean_iou: 0.5120WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
34/78 [============>.................] - ETA: 31:13 - loss: 0.4423 - accuracy: 0.9183 - mean_iou: 0.5128WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
35/78 [============>.................] - ETA: 30:32 - loss: 0.4437 - accuracy: 0.9179 - mean_iou: 0.5097WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
36/78 [============>.................] - ETA: 29:51 - loss: 0.4454 - accuracy: 0.9176 - mean_iou: 0.5078WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
37/78 [=============>................] - ETA: 29:10 - loss: 0.4456 - accuracy: 0.9173 - mean_iou: 0.5068WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
38/78 [=============>................] - ETA: 28:29 - loss: 0.4447 - accuracy: 0.9177 - mean_iou: 0.5094WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
39/78 [==============>...............] - ETA: 27:48 - loss: 0.4453 - accuracy: 0.9171 - mean_iou: 0.5107WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
40/78 [==============>...............] - ETA: 27:07 - loss: 0.4452 - accuracy: 0.9173 - mean_iou: 0.5100WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
41/78 [==============>...............] - ETA: 26:25 - loss: 0.4460 - accuracy: 0.9170 - mean_iou: 0.5098WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
42/78 [===============>..............] - ETA: 25:43 - loss: 0.4465 - accuracy: 0.9166 - mean_iou: 0.5067WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
43/78 [===============>..............] - ETA: 25:01 - loss: 0.4476 - accuracy: 0.9162 - mean_iou: 0.5051WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
44/78 [===============>..............] - ETA: 24:19 - loss: 0.4490 - accuracy: 0.9158 - mean_iou: 0.5038WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
45/78 [================>.............] - ETA: 23:36 - loss: 0.4492 - accuracy: 0.9158 - mean_iou: 0.5009WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
46/78 [================>.............] - ETA: 22:54 - loss: 0.4490 - accuracy: 0.9157 - mean_iou: 0.5008WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
47/78 [=================>............] - ETA: 22:12 - loss: 0.4483 - accuracy: 0.9160 - mean_iou: 0.5028WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
48/78 [=================>............] - ETA: 21:30 - loss: 0.4485 - accuracy: 0.9160 - mean_iou: 0.5034WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
49/78 [=================>............] - ETA: 20:47 - loss: 0.4481 - accuracy: 0.9163 - mean_iou: 0.5038WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
50/78 [==================>...........] - ETA: 20:06 - loss: 0.4476 - accuracy: 0.9167 - mean_iou: 0.5056WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
51/78 [==================>...........] - ETA: 19:23 - loss: 0.4493 - accuracy: 0.9165 - mean_iou: 0.5026WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
52/78 [===================>..........] - ETA: 18:41 - loss: 0.4488 - accuracy: 0.9168 - mean_iou: 0.5027WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
53/78 [===================>..........] - ETA: 17:58 - loss: 0.4491 - accuracy: 0.9168 - mean_iou: 0.5020WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
54/78 [===================>..........] - ETA: 17:15 - loss: 0.4501 - accuracy: 0.9165 - mean_iou: 0.5020WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
55/78 [====================>.........] - ETA: 16:32 - loss: 0.4508 - accuracy: 0.9163 - mean_iou: 0.5019WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
56/78 [====================>.........] - ETA: 15:49 - loss: 0.4507 - accuracy: 0.9161 - mean_iou: 0.5022WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
57/78 [====================>.........] - ETA: 15:06 - loss: 0.4504 - accuracy: 0.9160 - mean_iou: 0.5022WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
58/78 [=====================>........] - ETA: 14:23 - loss: 0.4507 - accuracy: 0.9160 - mean_iou: 0.5034WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
59/78 [=====================>........] - ETA: 13:40 - loss: 0.4505 - accuracy: 0.9160 - mean_iou: 0.5031WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
60/78 [======================>.......] - ETA: 12:57 - loss: 0.4506 - accuracy: 0.9158 - mean_iou: 0.5008WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
61/78 [======================>.......] - ETA: 12:14 - loss: 0.4505 - accuracy: 0.9157 - mean_iou: 0.5017WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
62/78 [======================>.......] - ETA: 11:31 - loss: 0.4505 - accuracy: 0.9155 - mean_iou: 0.5016WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
63/78 [=======================>......] - ETA: 10:48 - loss: 0.4498 - accuracy: 0.9156 - mean_iou: 0.5021WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
64/78 [=======================>......] - ETA: 10:05 - loss: 0.4492 - accuracy: 0.9157 - mean_iou: 0.5019WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
65/78 [========================>.....] - ETA: 9:22 - loss: 0.4489 - accuracy: 0.9157 - mean_iou: 0.5019 WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
66/78 [========================>.....] - ETA: 8:39 - loss: 0.4492 - accuracy: 0.9157 - mean_iou: 0.5026WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
67/78 [========================>.....] - ETA: 7:56 - loss: 0.4491 - accuracy: 0.9157 - mean_iou: 0.5024WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
68/78 [=========================>....] - ETA: 7:12 - loss: 0.4493 - accuracy: 0.9155 - mean_iou: 0.5020WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
69/78 [=========================>....] - ETA: 6:29 - loss: 0.4491 - accuracy: 0.9154 - mean_iou: 0.5008WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
70/78 [=========================>....] - ETA: 5:45 - loss: 0.4488 - accuracy: 0.9156 - mean_iou: 0.4999WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
71/78 [==========================>...] - ETA: 5:02 - loss: 0.4487 - accuracy: 0.9157 - mean_iou: 0.5005WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
72/78 [==========================>...] - ETA: 4:18 - loss: 0.4491 - accuracy: 0.9158 - mean_iou: 0.4993WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
73/78 [===========================>..] - ETA: 3:35 - loss: 0.4488 - accuracy: 0.9160 - mean_iou: 0.5015WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
74/78 [===========================>..] - ETA: 2:52 - loss: 0.4478 - accuracy: 0.9162 - mean_iou: 0.5035WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
75/78 [===========================>..] - ETA: 2:09 - loss: 0.4475 - accuracy: 0.9164 - mean_iou: 0.5033WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
76/78 [============================>.] - ETA: 1:26 - loss: 0.4484 - accuracy: 0.9164 - mean_iou: 0.5034WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
77/78 [============================>.] - ETA: 42s - loss: 0.4488 - accuracy: 0.9163 - mean_iou: 0.5031 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
78/78 [==============================] - ETA: 0s - loss: 0.4485 - accuracy: 0.9164 - mean_iou: 0.5042 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
78/78 [==============================] - 3636s 47s/step - loss: 0.4485 - accuracy: 0.9164 - mean_iou: 0.5042 - val_loss: 0.4465 - val_accuracy: 0.9280 - val_mean_iou: 0.5568 - lr: 9.8429e-04
Epoch 4/6
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 1/78 [..............................] - ETA: 0s - loss: 0.4138 - accuracy: 0.9200 - mean_iou: 0.5817WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 2/78 [..............................] - ETA: 28:21 - loss: 0.4127 - accuracy: 0.9305 - mean_iou: 0.5895WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 3/78 [>.............................] - ETA: 36:56 - loss: 0.4155 - accuracy: 0.9296 - mean_iou: 0.5613WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 4/78 [>.............................] - ETA: 40:56 - loss: 0.4246 - accuracy: 0.9273 - mean_iou: 0.5392WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 5/78 [>.............................] - ETA: 42:52 - loss: 0.4288 - accuracy: 0.9259 - mean_iou: 0.5367WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 6/78 [=>............................] - ETA: 44:02 - loss: 0.4215 - accuracy: 0.9275 - mean_iou: 0.5343WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 7/78 [=>............................] - ETA: 44:37 - loss: 0.4292 - accuracy: 0.9282 - mean_iou: 0.5318WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 8/78 [==>...........................] - ETA: 44:49 - loss: 0.4314 - accuracy: 0.9279 - mean_iou: 0.5224WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 9/78 [==>...........................] - ETA: 44:52 - loss: 0.4341 - accuracy: 0.9260 - mean_iou: 0.5307WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
10/78 [==>...........................] - ETA: 44:43 - loss: 0.4339 - accuracy: 0.9254 - mean_iou: 0.5269WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
11/78 [===>..........................] - ETA: 44:32 - loss: 0.4376 - accuracy: 0.9235 - mean_iou: 0.5162WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
12/78 [===>..........................] - ETA: 44:16 - loss: 0.4388 - accuracy: 0.9238 - mean_iou: 0.5188WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
13/78 [====>.........................] - ETA: 43:53 - loss: 0.4336 - accuracy: 0.9244 - mean_iou: 0.5160WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
14/78 [====>.........................] - ETA: 44:09 - loss: 0.4376 - accuracy: 0.9235 - mean_iou: 0.5053WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
15/78 [====>.........................] - ETA: 43:38 - loss: 0.4352 - accuracy: 0.9236 - mean_iou: 0.4991WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
16/78 [=====>........................] - ETA: 43:05 - loss: 0.4320 - accuracy: 0.9242 - mean_iou: 0.5051WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
17/78 [=====>........................] - ETA: 42:31 - loss: 0.4333 - accuracy: 0.9232 - mean_iou: 0.5098WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
18/78 [=====>........................] - ETA: 41:54 - loss: 0.4364 - accuracy: 0.9228 - mean_iou: 0.5039WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
19/78 [======>.......................] - ETA: 41:19 - loss: 0.4353 - accuracy: 0.9231 - mean_iou: 0.5082WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
20/78 [======>.......................] - ETA: 40:41 - loss: 0.4356 - accuracy: 0.9227 - mean_iou: 0.5074WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
21/78 [=======>......................] - ETA: 40:05 - loss: 0.4374 - accuracy: 0.9221 - mean_iou: 0.4987WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
22/78 [=======>......................] - ETA: 39:27 - loss: 0.4395 - accuracy: 0.9219 - mean_iou: 0.4967WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
23/78 [=======>......................] - ETA: 38:48 - loss: 0.4381 - accuracy: 0.9227 - mean_iou: 0.4976WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
24/78 [========>.....................] - ETA: 38:09 - loss: 0.4370 - accuracy: 0.9227 - mean_iou: 0.4993WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
25/78 [========>.....................] - ETA: 37:29 - loss: 0.4359 - accuracy: 0.9229 - mean_iou: 0.5067WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
26/78 [=========>....................] - ETA: 36:50 - loss: 0.4370 - accuracy: 0.9223 - mean_iou: 0.5084WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
27/78 [=========>....................] - ETA: 36:10 - loss: 0.4370 - accuracy: 0.9226 - mean_iou: 0.5163WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
28/78 [=========>....................] - ETA: 35:29 - loss: 0.4383 - accuracy: 0.9215 - mean_iou: 0.5157WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
29/78 [==========>...................] - ETA: 34:49 - loss: 0.4387 - accuracy: 0.9209 - mean_iou: 0.5138WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
30/78 [==========>...................] - ETA: 34:09 - loss: 0.4377 - accuracy: 0.9205 - mean_iou: 0.5139WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
31/78 [==========>...................] - ETA: 33:29 - loss: 0.4383 - accuracy: 0.9199 - mean_iou: 0.5117WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
32/78 [===========>..................] - ETA: 32:48 - loss: 0.4388 - accuracy: 0.9196 - mean_iou: 0.5121WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
33/78 [===========>..................] - ETA: 32:09 - loss: 0.4394 - accuracy: 0.9191 - mean_iou: 0.5125WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
34/78 [============>.................] - ETA: 31:29 - loss: 0.4400 - accuracy: 0.9187 - mean_iou: 0.5109WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
35/78 [============>.................] - ETA: 30:48 - loss: 0.4402 - accuracy: 0.9183 - mean_iou: 0.5106WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
36/78 [============>.................] - ETA: 30:07 - loss: 0.4402 - accuracy: 0.9181 - mean_iou: 0.5094WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
37/78 [=============>................] - ETA: 29:25 - loss: 0.4403 - accuracy: 0.9183 - mean_iou: 0.5097WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
38/78 [=============>................] - ETA: 28:43 - loss: 0.4406 - accuracy: 0.9185 - mean_iou: 0.5104WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
39/78 [==============>...............] - ETA: 28:01 - loss: 0.4412 - accuracy: 0.9187 - mean_iou: 0.5073WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
40/78 [==============>...............] - ETA: 27:19 - loss: 0.4412 - accuracy: 0.9190 - mean_iou: 0.5056WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
41/78 [==============>...............] - ETA: 26:37 - loss: 0.4419 - accuracy: 0.9187 - mean_iou: 0.5048WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
42/78 [===============>..............] - ETA: 25:54 - loss: 0.4417 - accuracy: 0.9187 - mean_iou: 0.5035WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
43/78 [===============>..............] - ETA: 25:12 - loss: 0.4411 - accuracy: 0.9189 - mean_iou: 0.5055WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
44/78 [===============>..............] - ETA: 24:29 - loss: 0.4400 - accuracy: 0.9192 - mean_iou: 0.5076WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
45/78 [================>.............] - ETA: 23:46 - loss: 0.4396 - accuracy: 0.9192 - mean_iou: 0.5094WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
46/78 [================>.............] - ETA: 23:05 - loss: 0.4392 - accuracy: 0.9192 - mean_iou: 0.5105WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
47/78 [=================>............] - ETA: 22:22 - loss: 0.4392 - accuracy: 0.9191 - mean_iou: 0.5100WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
48/78 [=================>............] - ETA: 21:39 - loss: 0.4399 - accuracy: 0.9188 - mean_iou: 0.5090WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
49/78 [=================>............] - ETA: 20:57 - loss: 0.4399 - accuracy: 0.9188 - mean_iou: 0.5086WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
50/78 [==================>...........] - ETA: 20:14 - loss: 0.4406 - accuracy: 0.9189 - mean_iou: 0.5079WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
51/78 [==================>...........] - ETA: 19:31 - loss: 0.4407 - accuracy: 0.9188 - mean_iou: 0.5067WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
52/78 [===================>..........] - ETA: 18:48 - loss: 0.4409 - accuracy: 0.9189 - mean_iou: 0.5064WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
53/78 [===================>..........] - ETA: 18:05 - loss: 0.4403 - accuracy: 0.9193 - mean_iou: 0.5094WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
54/78 [===================>..........] - ETA: 17:21 - loss: 0.4408 - accuracy: 0.9192 - mean_iou: 0.5103WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
55/78 [====================>.........] - ETA: 16:38 - loss: 0.4400 - accuracy: 0.9197 - mean_iou: 0.5133WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
56/78 [====================>.........] - ETA: 15:55 - loss: 0.4405 - accuracy: 0.9199 - mean_iou: 0.5131WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
57/78 [====================>.........] - ETA: 15:11 - loss: 0.4401 - accuracy: 0.9201 - mean_iou: 0.5145WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
58/78 [=====================>........] - ETA: 14:28 - loss: 0.4398 - accuracy: 0.9203 - mean_iou: 0.5147WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
59/78 [=====================>........] - ETA: 13:45 - loss: 0.4393 - accuracy: 0.9204 - mean_iou: 0.5141WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
60/78 [======================>.......] - ETA: 13:01 - loss: 0.4395 - accuracy: 0.9203 - mean_iou: 0.5152WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
61/78 [======================>.......] - ETA: 12:18 - loss: 0.4399 - accuracy: 0.9202 - mean_iou: 0.5149WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
62/78 [======================>.......] - ETA: 11:35 - loss: 0.4407 - accuracy: 0.9203 - mean_iou: 0.5132WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
63/78 [=======================>......] - ETA: 10:51 - loss: 0.4412 - accuracy: 0.9202 - mean_iou: 0.5134WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
64/78 [=======================>......] - ETA: 10:08 - loss: 0.4413 - accuracy: 0.9203 - mean_iou: 0.5130WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
65/78 [========================>.....] - ETA: 9:25 - loss: 0.4419 - accuracy: 0.9201 - mean_iou: 0.5118 WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
66/78 [========================>.....] - ETA: 8:41 - loss: 0.4422 - accuracy: 0.9202 - mean_iou: 0.5113WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
67/78 [========================>.....] - ETA: 7:58 - loss: 0.4431 - accuracy: 0.9200 - mean_iou: 0.5098WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
68/78 [=========================>....] - ETA: 7:14 - loss: 0.4429 - accuracy: 0.9200 - mean_iou: 0.5095WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
69/78 [=========================>....] - ETA: 6:30 - loss: 0.4437 - accuracy: 0.9198 - mean_iou: 0.5089WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
70/78 [=========================>....] - ETA: 5:47 - loss: 0.4440 - accuracy: 0.9197 - mean_iou: 0.5094WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
71/78 [==========================>...] - ETA: 5:03 - loss: 0.4437 - accuracy: 0.9198 - mean_iou: 0.5093WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
72/78 [==========================>...] - ETA: 4:19 - loss: 0.4433 - accuracy: 0.9199 - mean_iou: 0.5093WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
73/78 [===========================>..] - ETA: 3:36 - loss: 0.4429 - accuracy: 0.9200 - mean_iou: 0.5091WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
74/78 [===========================>..] - ETA: 2:52 - loss: 0.4428 - accuracy: 0.9199 - mean_iou: 0.5086WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
75/78 [===========================>..] - ETA: 2:09 - loss: 0.4431 - accuracy: 0.9197 - mean_iou: 0.5079WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
76/78 [============================>.] - ETA: 1:26 - loss: 0.4426 - accuracy: 0.9198 - mean_iou: 0.5085WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
77/78 [============================>.] - ETA: 43s - loss: 0.4427 - accuracy: 0.9196 - mean_iou: 0.5081 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
78/78 [==============================] - ETA: 0s - loss: 0.4418 - accuracy: 0.9197 - mean_iou: 0.5098 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
78/78 [==============================] - 3641s 47s/step - loss: 0.4418 - accuracy: 0.9197 - mean_iou: 0.5098 - val_loss: 0.4483 - val_accuracy: 0.9212 - val_mean_iou: 0.5062 - lr: 9.6489e-04
Epoch 5/6
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 1/78 [..............................] - ETA: 0s - loss: 0.3991 - accuracy: 0.9185 - mean_iou: 0.5028WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 2/78 [..............................] - ETA: 28:41 - loss: 0.4081 - accuracy: 0.9182 - mean_iou: 0.5235WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 3/78 [>.............................] - ETA: 37:06 - loss: 0.4184 - accuracy: 0.9203 - mean_iou: 0.5050WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 4/78 [>.............................] - ETA: 40:55 - loss: 0.4191 - accuracy: 0.9227 - mean_iou: 0.5336WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 5/78 [>.............................] - ETA: 42:58 - loss: 0.4216 - accuracy: 0.9217 - mean_iou: 0.5419WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 6/78 [=>............................] - ETA: 44:09 - loss: 0.4159 - accuracy: 0.9231 - mean_iou: 0.5470WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 7/78 [=>............................] - ETA: 44:44 - loss: 0.4164 - accuracy: 0.9222 - mean_iou: 0.5367WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 8/78 [==>...........................] - ETA: 44:57 - loss: 0.4135 - accuracy: 0.9227 - mean_iou: 0.5348WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 9/78 [==>...........................] - ETA: 45:02 - loss: 0.4157 - accuracy: 0.9238 - mean_iou: 0.5156WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
10/78 [==>...........................] - ETA: 44:54 - loss: 0.4159 - accuracy: 0.9242 - mean_iou: 0.5212WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
11/78 [===>..........................] - ETA: 44:40 - loss: 0.4188 - accuracy: 0.9248 - mean_iou: 0.5267WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
12/78 [===>..........................] - ETA: 44:20 - loss: 0.4182 - accuracy: 0.9252 - mean_iou: 0.5290WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
13/78 [====>.........................] - ETA: 43:57 - loss: 0.4172 - accuracy: 0.9260 - mean_iou: 0.5293WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
14/78 [====>.........................] - ETA: 43:32 - loss: 0.4215 - accuracy: 0.9260 - mean_iou: 0.5273WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
15/78 [====>.........................] - ETA: 43:02 - loss: 0.4246 - accuracy: 0.9250 - mean_iou: 0.5266WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
16/78 [=====>........................] - ETA: 42:32 - loss: 0.4228 - accuracy: 0.9266 - mean_iou: 0.5335WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
17/78 [=====>........................] - ETA: 42:00 - loss: 0.4261 - accuracy: 0.9260 - mean_iou: 0.5322WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
18/78 [=====>........................] - ETA: 41:28 - loss: 0.4229 - accuracy: 0.9262 - mean_iou: 0.5370WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
19/78 [======>.......................] - ETA: 40:54 - loss: 0.4239 - accuracy: 0.9265 - mean_iou: 0.5331WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
20/78 [======>.......................] - ETA: 40:18 - loss: 0.4265 - accuracy: 0.9253 - mean_iou: 0.5317WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
21/78 [=======>......................] - ETA: 39:42 - loss: 0.4289 - accuracy: 0.9250 - mean_iou: 0.5273WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
22/78 [=======>......................] - ETA: 39:05 - loss: 0.4280 - accuracy: 0.9248 - mean_iou: 0.5234WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
23/78 [=======>......................] - ETA: 38:29 - loss: 0.4264 - accuracy: 0.9248 - mean_iou: 0.5234WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
24/78 [========>.....................] - ETA: 37:52 - loss: 0.4275 - accuracy: 0.9247 - mean_iou: 0.5193WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
25/78 [========>.....................] - ETA: 37:14 - loss: 0.4277 - accuracy: 0.9240 - mean_iou: 0.5181WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
26/78 [=========>....................] - ETA: 36:36 - loss: 0.4253 - accuracy: 0.9246 - mean_iou: 0.5212WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
27/78 [=========>....................] - ETA: 35:59 - loss: 0.4237 - accuracy: 0.9246 - mean_iou: 0.5249WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
28/78 [=========>....................] - ETA: 35:20 - loss: 0.4252 - accuracy: 0.9241 - mean_iou: 0.5195WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
29/78 [==========>...................] - ETA: 34:40 - loss: 0.4257 - accuracy: 0.9238 - mean_iou: 0.5157WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
30/78 [==========>...................] - ETA: 34:01 - loss: 0.4239 - accuracy: 0.9243 - mean_iou: 0.5222WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
31/78 [==========>...................] - ETA: 33:21 - loss: 0.4228 - accuracy: 0.9244 - mean_iou: 0.5202WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
32/78 [===========>..................] - ETA: 32:41 - loss: 0.4276 - accuracy: 0.9231 - mean_iou: 0.5164WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
33/78 [===========>..................] - ETA: 32:00 - loss: 0.4266 - accuracy: 0.9233 - mean_iou: 0.5185WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
34/78 [============>.................] - ETA: 31:19 - loss: 0.4267 - accuracy: 0.9232 - mean_iou: 0.5213WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
35/78 [============>.................] - ETA: 30:38 - loss: 0.4261 - accuracy: 0.9235 - mean_iou: 0.5217WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
36/78 [============>.................] - ETA: 29:57 - loss: 0.4264 - accuracy: 0.9234 - mean_iou: 0.5208WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
37/78 [=============>................] - ETA: 29:16 - loss: 0.4249 - accuracy: 0.9239 - mean_iou: 0.5236WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
38/78 [=============>................] - ETA: 28:34 - loss: 0.4246 - accuracy: 0.9240 - mean_iou: 0.5235WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
39/78 [==============>...............] - ETA: 27:52 - loss: 0.4255 - accuracy: 0.9240 - mean_iou: 0.5231WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
40/78 [==============>...............] - ETA: 27:10 - loss: 0.4263 - accuracy: 0.9241 - mean_iou: 0.5238WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
41/78 [==============>...............] - ETA: 26:29 - loss: 0.4265 - accuracy: 0.9242 - mean_iou: 0.5250WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
42/78 [===============>..............] - ETA: 25:46 - loss: 0.4260 - accuracy: 0.9244 - mean_iou: 0.5249WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
43/78 [===============>..............] - ETA: 25:04 - loss: 0.4260 - accuracy: 0.9247 - mean_iou: 0.5237WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
44/78 [===============>..............] - ETA: 24:22 - loss: 0.4258 - accuracy: 0.9247 - mean_iou: 0.5228WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
45/78 [================>.............] - ETA: 23:40 - loss: 0.4278 - accuracy: 0.9239 - mean_iou: 0.5210WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
46/78 [================>.............] - ETA: 22:58 - loss: 0.4275 - accuracy: 0.9239 - mean_iou: 0.5224WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
47/78 [=================>............] - ETA: 22:15 - loss: 0.4279 - accuracy: 0.9238 - mean_iou: 0.5210WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
48/78 [=================>............] - ETA: 21:32 - loss: 0.4270 - accuracy: 0.9238 - mean_iou: 0.5226WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
49/78 [=================>............] - ETA: 20:50 - loss: 0.4263 - accuracy: 0.9238 - mean_iou: 0.5242WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
50/78 [==================>...........] - ETA: 20:07 - loss: 0.4262 - accuracy: 0.9238 - mean_iou: 0.5237WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
51/78 [==================>...........] - ETA: 19:25 - loss: 0.4274 - accuracy: 0.9232 - mean_iou: 0.5214WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
52/78 [===================>..........] - ETA: 18:42 - loss: 0.4271 - accuracy: 0.9230 - mean_iou: 0.5211WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
53/78 [===================>..........] - ETA: 17:59 - loss: 0.4264 - accuracy: 0.9229 - mean_iou: 0.5214WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
54/78 [===================>..........] - ETA: 17:16 - loss: 0.4263 - accuracy: 0.9228 - mean_iou: 0.5218WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
55/78 [====================>.........] - ETA: 16:33 - loss: 0.4260 - accuracy: 0.9229 - mean_iou: 0.5205WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
56/78 [====================>.........] - ETA: 15:50 - loss: 0.4260 - accuracy: 0.9230 - mean_iou: 0.5213WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
57/78 [====================>.........] - ETA: 15:07 - loss: 0.4258 - accuracy: 0.9229 - mean_iou: 0.5215WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
58/78 [=====================>........] - ETA: 14:24 - loss: 0.4250 - accuracy: 0.9231 - mean_iou: 0.5237WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
59/78 [=====================>........] - ETA: 13:41 - loss: 0.4247 - accuracy: 0.9232 - mean_iou: 0.5218WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
60/78 [======================>.......] - ETA: 12:58 - loss: 0.4245 - accuracy: 0.9233 - mean_iou: 0.5233WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
61/78 [======================>.......] - ETA: 12:15 - loss: 0.4243 - accuracy: 0.9235 - mean_iou: 0.5253WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
62/78 [======================>.......] - ETA: 11:32 - loss: 0.4233 - accuracy: 0.9240 - mean_iou: 0.5281WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
63/78 [=======================>......] - ETA: 10:49 - loss: 0.4234 - accuracy: 0.9241 - mean_iou: 0.5293WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
64/78 [=======================>......] - ETA: 10:06 - loss: 0.4227 - accuracy: 0.9243 - mean_iou: 0.5294WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
65/78 [========================>.....] - ETA: 9:23 - loss: 0.4218 - accuracy: 0.9245 - mean_iou: 0.5307 WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
66/78 [========================>.....] - ETA: 8:40 - loss: 0.4221 - accuracy: 0.9243 - mean_iou: 0.5312WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
67/78 [========================>.....] - ETA: 7:56 - loss: 0.4220 - accuracy: 0.9243 - mean_iou: 0.5315WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
68/78 [=========================>....] - ETA: 7:13 - loss: 0.4221 - accuracy: 0.9241 - mean_iou: 0.5313WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
69/78 [=========================>....] - ETA: 6:29 - loss: 0.4217 - accuracy: 0.9241 - mean_iou: 0.5322WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
70/78 [=========================>....] - ETA: 5:45 - loss: 0.4218 - accuracy: 0.9239 - mean_iou: 0.5314WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
71/78 [==========================>...] - ETA: 5:02 - loss: 0.4215 - accuracy: 0.9240 - mean_iou: 0.5321WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
72/78 [==========================>...] - ETA: 4:18 - loss: 0.4218 - accuracy: 0.9239 - mean_iou: 0.5320WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
73/78 [===========================>..] - ETA: 3:35 - loss: 0.4215 - accuracy: 0.9239 - mean_iou: 0.5314WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
74/78 [===========================>..] - ETA: 2:52 - loss: 0.4226 - accuracy: 0.9239 - mean_iou: 0.5301WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
75/78 [===========================>..] - ETA: 2:09 - loss: 0.4231 - accuracy: 0.9238 - mean_iou: 0.5297WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
76/78 [============================>.] - ETA: 1:25 - loss: 0.4235 - accuracy: 0.9237 - mean_iou: 0.5312WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
77/78 [============================>.] - ETA: 42s - loss: 0.4238 - accuracy: 0.9239 - mean_iou: 0.5307 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
78/78 [==============================] - ETA: 0s - loss: 0.4233 - accuracy: 0.9241 - mean_iou: 0.5322 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
78/78 [==============================] - 3627s 47s/step - loss: 0.4233 - accuracy: 0.9241 - mean_iou: 0.5322 - val_loss: 0.4118 - val_accuracy: 0.9300 - val_mean_iou: 0.5509 - lr: 9.3815e-04
Epoch 6/6
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 1/78 [..............................] - ETA: 0s - loss: 0.4062 - accuracy: 0.9323 - mean_iou: 0.5790WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 2/78 [..............................] - ETA: 28:45 - loss: 0.3978 - accuracy: 0.9370 - mean_iou: 0.5877WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 3/78 [>.............................] - ETA: 37:04 - loss: 0.3997 - accuracy: 0.9359 - mean_iou: 0.5492WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 4/78 [>.............................] - ETA: 41:45 - loss: 0.4152 - accuracy: 0.9300 - mean_iou: 0.5340WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 5/78 [>.............................] - ETA: 43:33 - loss: 0.4207 - accuracy: 0.9251 - mean_iou: 0.5312WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 6/78 [=>............................] - ETA: 44:32 - loss: 0.4155 - accuracy: 0.9271 - mean_iou: 0.5293WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 7/78 [=>............................] - ETA: 45:01 - loss: 0.4110 - accuracy: 0.9275 - mean_iou: 0.5282WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 8/78 [==>...........................] - ETA: 45:12 - loss: 0.4111 - accuracy: 0.9282 - mean_iou: 0.5368WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
 9/78 [==>...........................] - ETA: 45:12 - loss: 0.4129 - accuracy: 0.9267 - mean_iou: 0.5329WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
10/78 [==>...........................] - ETA: 45:05 - loss: 0.4157 - accuracy: 0.9262 - mean_iou: 0.5456WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
11/78 [===>..........................] - ETA: 44:52 - loss: 0.4192 - accuracy: 0.9254 - mean_iou: 0.5395WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
12/78 [===>..........................] - ETA: 44:32 - loss: 0.4235 - accuracy: 0.9248 - mean_iou: 0.5284WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
13/78 [====>.........................] - ETA: 44:12 - loss: 0.4230 - accuracy: 0.9252 - mean_iou: 0.5324WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
14/78 [====>.........................] - ETA: 43:46 - loss: 0.4284 - accuracy: 0.9237 - mean_iou: 0.5205WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
15/78 [====>.........................] - ETA: 43:17 - loss: 0.4291 - accuracy: 0.9238 - mean_iou: 0.5186WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
16/78 [=====>........................] - ETA: 42:47 - loss: 0.4250 - accuracy: 0.9249 - mean_iou: 0.5235WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
17/78 [=====>........................] - ETA: 42:15 - loss: 0.4255 - accuracy: 0.9249 - mean_iou: 0.5261WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
18/78 [=====>........................] - ETA: 41:40 - loss: 0.4222 - accuracy: 0.9259 - mean_iou: 0.5320WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
19/78 [======>.......................] - ETA: 41:05 - loss: 0.4220 - accuracy: 0.9254 - mean_iou: 0.5327WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
20/78 [======>.......................] - ETA: 40:28 - loss: 0.4231 - accuracy: 0.9246 - mean_iou: 0.5281WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
21/78 [=======>......................] - ETA: 39:52 - loss: 0.4230 - accuracy: 0.9249 - mean_iou: 0.5271WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
22/78 [=======>......................] - ETA: 39:15 - loss: 0.4249 - accuracy: 0.9245 - mean_iou: 0.5253WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
23/78 [=======>......................] - ETA: 38:37 - loss: 0.4237 - accuracy: 0.9248 - mean_iou: 0.5261WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
24/78 [========>.....................] - ETA: 37:58 - loss: 0.4251 - accuracy: 0.9241 - mean_iou: 0.5275WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
25/78 [========>.....................] - ETA: 37:19 - loss: 0.4254 - accuracy: 0.9236 - mean_iou: 0.5251WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
26/78 [=========>....................] - ETA: 36:40 - loss: 0.4268 - accuracy: 0.9232 - mean_iou: 0.5241WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
27/78 [=========>....................] - ETA: 36:00 - loss: 0.4247 - accuracy: 0.9237 - mean_iou: 0.5295WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
28/78 [=========>....................] - ETA: 35:20 - loss: 0.4249 - accuracy: 0.9240 - mean_iou: 0.5299WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
29/78 [==========>...................] - ETA: 34:40 - loss: 0.4253 - accuracy: 0.9238 - mean_iou: 0.5262WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
30/78 [==========>...................] - ETA: 34:00 - loss: 0.4264 - accuracy: 0.9235 - mean_iou: 0.5238WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
31/78 [==========>...................] - ETA: 33:19 - loss: 0.4245 - accuracy: 0.9238 - mean_iou: 0.5260WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
32/78 [===========>..................] - ETA: 32:38 - loss: 0.4243 - accuracy: 0.9238 - mean_iou: 0.5243WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
33/78 [===========>..................] - ETA: 31:57 - loss: 0.4238 - accuracy: 0.9237 - mean_iou: 0.5217WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
34/78 [============>.................] - ETA: 31:19 - loss: 0.4232 - accuracy: 0.9237 - mean_iou: 0.5220WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
35/78 [============>.................] - ETA: 30:38 - loss: 0.4228 - accuracy: 0.9235 - mean_iou: 0.5210WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
36/78 [============>.................] - ETA: 29:56 - loss: 0.4228 - accuracy: 0.9235 - mean_iou: 0.5212WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
37/78 [=============>................] - ETA: 29:14 - loss: 0.4237 - accuracy: 0.9235 - mean_iou: 0.5176WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
38/78 [=============>................] - ETA: 28:32 - loss: 0.4235 - accuracy: 0.9231 - mean_iou: 0.5176WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
39/78 [==============>...............] - ETA: 27:50 - loss: 0.4233 - accuracy: 0.9231 - mean_iou: 0.5182WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
40/78 [==============>...............] - ETA: 27:08 - loss: 0.4218 - accuracy: 0.9235 - mean_iou: 0.5195WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
41/78 [==============>...............] - ETA: 26:26 - loss: 0.4223 - accuracy: 0.9236 - mean_iou: 0.5180WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
42/78 [===============>..............] - ETA: 25:45 - loss: 0.4237 - accuracy: 0.9233 - mean_iou: 0.5170WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
43/78 [===============>..............] - ETA: 25:03 - loss: 0.4244 - accuracy: 0.9229 - mean_iou: 0.5148WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
44/78 [===============>..............] - ETA: 24:21 - loss: 0.4239 - accuracy: 0.9233 - mean_iou: 0.5179WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
45/78 [================>.............] - ETA: 23:39 - loss: 0.4230 - accuracy: 0.9236 - mean_iou: 0.5205WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
46/78 [================>.............] - ETA: 22:56 - loss: 0.4228 - accuracy: 0.9238 - mean_iou: 0.5223WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
47/78 [=================>............] - ETA: 22:13 - loss: 0.4236 - accuracy: 0.9236 - mean_iou: 0.5204WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
48/78 [=================>............] - ETA: 21:31 - loss: 0.4229 - accuracy: 0.9237 - mean_iou: 0.5223WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
49/78 [=================>............] - ETA: 20:48 - loss: 0.4232 - accuracy: 0.9235 - mean_iou: 0.5239WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
50/78 [==================>...........] - ETA: 20:06 - loss: 0.4218 - accuracy: 0.9238 - mean_iou: 0.5261WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
51/78 [==================>...........] - ETA: 19:23 - loss: 0.4211 - accuracy: 0.9238 - mean_iou: 0.5278WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
52/78 [===================>..........] - ETA: 18:40 - loss: 0.4218 - accuracy: 0.9236 - mean_iou: 0.5260WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
53/78 [===================>..........] - ETA: 17:57 - loss: 0.4220 - accuracy: 0.9233 - mean_iou: 0.5250WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
54/78 [===================>..........] - ETA: 17:15 - loss: 0.4215 - accuracy: 0.9233 - mean_iou: 0.5259WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
55/78 [====================>.........] - ETA: 16:32 - loss: 0.4214 - accuracy: 0.9232 - mean_iou: 0.5251WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
56/78 [====================>.........] - ETA: 15:49 - loss: 0.4214 - accuracy: 0.9231 - mean_iou: 0.5244WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
57/78 [====================>.........] - ETA: 15:06 - loss: 0.4211 - accuracy: 0.9230 - mean_iou: 0.5249WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
58/78 [=====================>........] - ETA: 14:23 - loss: 0.4204 - accuracy: 0.9231 - mean_iou: 0.5269WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
59/78 [=====================>........] - ETA: 13:40 - loss: 0.4200 - accuracy: 0.9232 - mean_iou: 0.5271WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
60/78 [======================>.......] - ETA: 12:57 - loss: 0.4194 - accuracy: 0.9234 - mean_iou: 0.5268WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
61/78 [======================>.......] - ETA: 12:14 - loss: 0.4189 - accuracy: 0.9235 - mean_iou: 0.5264WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
62/78 [======================>.......] - ETA: 11:31 - loss: 0.4183 - accuracy: 0.9236 - mean_iou: 0.5263WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
63/78 [=======================>......] - ETA: 10:48 - loss: 0.4184 - accuracy: 0.9234 - mean_iou: 0.5269WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
64/78 [=======================>......] - ETA: 10:05 - loss: 0.4178 - accuracy: 0.9238 - mean_iou: 0.5278WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
65/78 [========================>.....] - ETA: 9:22 - loss: 0.4185 - accuracy: 0.9237 - mean_iou: 0.5270 WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
66/78 [========================>.....] - ETA: 8:39 - loss: 0.4182 - accuracy: 0.9239 - mean_iou: 0.5268WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
67/78 [========================>.....] - ETA: 7:56 - loss: 0.4176 - accuracy: 0.9241 - mean_iou: 0.5288WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
68/78 [=========================>....] - ETA: 7:12 - loss: 0.4182 - accuracy: 0.9240 - mean_iou: 0.5300WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
69/78 [=========================>....] - ETA: 6:29 - loss: 0.4170 - accuracy: 0.9245 - mean_iou: 0.5329WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
70/78 [=========================>....] - ETA: 5:45 - loss: 0.4159 - accuracy: 0.9249 - mean_iou: 0.5355WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
71/78 [==========================>...] - ETA: 5:02 - loss: 0.4154 - accuracy: 0.9250 - mean_iou: 0.5363WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
72/78 [==========================>...] - ETA: 4:18 - loss: 0.4150 - accuracy: 0.9253 - mean_iou: 0.5370WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
73/78 [===========================>..] - ETA: 3:35 - loss: 0.4144 - accuracy: 0.9256 - mean_iou: 0.5364WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
74/78 [===========================>..] - ETA: 2:52 - loss: 0.4133 - accuracy: 0.9259 - mean_iou: 0.5386WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
75/78 [===========================>..] - ETA: 2:09 - loss: 0.4131 - accuracy: 0.9262 - mean_iou: 0.5397WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
76/78 [============================>.] - ETA: 1:26 - loss: 0.4123 - accuracy: 0.9264 - mean_iou: 0.5403WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
77/78 [============================>.] - ETA: 42s - loss: 0.4130 - accuracy: 0.9262 - mean_iou: 0.5408 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:Can save best model only with val_accuracy available, skipping.
78/78 [==============================] - ETA: 0s - loss: 0.4127 - accuracy: 0.9264 - mean_iou: 0.5418 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
78/78 [==============================] - 3629s 47s/step - loss: 0.4127 - accuracy: 0.9264 - mean_iou: 0.5418 - val_loss: 0.4492 - val_accuracy: 0.9331 - val_mean_iou: 0.5890 - lr: 9.0451e-04

model.load_weights("4000_checkpoint.hdf5")

from tensorflow.keras.callbacks import ModelCheckpoint
checkpoint_file="4000_checkpoint_1.hdf5"
checkpoint = ModelCheckpoint(checkpoint_file,monitor='accuracy',verbose=1,save_best_only=True,mode='max',save_freq=1)

history = model.fit_generator(train_gen, validation_data=valid_gen, callbacks=[learning_rate,checkpoint], epochs=6, workers=8, use_multiprocessing=True)

WARNING:tensorflow:From <ipython-input-50-af21aa0f54e4>:1: Model.fit_generator (from tensorflow.python.keras.engine.training) is deprecated and will be removed in a future version.
Instructions for updating:
Please use Model.fit, which supports generators.
Epoch 1/6
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.

Epoch 00001: accuracy improved from -inf to 0.94644, saving model to 4000_checkpoint_1.hdf5
 1/78 [..............................] - ETA: 0s - loss: 0.3660 - accuracy: 0.9464 - mean_iou: 0.5925
Epoch 00001: accuracy did not improve from 0.94644
 2/78 [..............................] - ETA: 32:15 - loss: 0.4199 - accuracy: 0.9385 - mean_iou: 0.5853
Epoch 00001: accuracy did not improve from 0.94644
 3/78 [>.............................] - ETA: 41:55 - loss: 0.4499 - accuracy: 0.9325 - mean_iou: 0.5496
Epoch 00001: accuracy did not improve from 0.94644
 4/78 [>.............................] - ETA: 46:25 - loss: 0.4523 - accuracy: 0.9316 - mean_iou: 0.5151
Epoch 00001: accuracy did not improve from 0.94644
 5/78 [>.............................] - ETA: 48:44 - loss: 0.4461 - accuracy: 0.9328 - mean_iou: 0.4989
Epoch 00001: accuracy did not improve from 0.94644
 6/78 [=>............................] - ETA: 50:01 - loss: 0.4487 - accuracy: 0.9305 - mean_iou: 0.4751
Epoch 00001: accuracy did not improve from 0.94644
 7/78 [=>............................] - ETA: 50:43 - loss: 0.4398 - accuracy: 0.9321 - mean_iou: 0.4881
Epoch 00001: accuracy did not improve from 0.94644
 8/78 [==>...........................] - ETA: 51:14 - loss: 0.4404 - accuracy: 0.9307 - mean_iou: 0.4896
Epoch 00001: accuracy did not improve from 0.94644
 9/78 [==>...........................] - ETA: 51:16 - loss: 0.4394 - accuracy: 0.9310 - mean_iou: 0.4911
Epoch 00001: accuracy did not improve from 0.94644
10/78 [==>...........................] - ETA: 51:07 - loss: 0.4404 - accuracy: 0.9287 - mean_iou: 0.4955
Epoch 00001: accuracy did not improve from 0.94644
11/78 [===>..........................] - ETA: 50:55 - loss: 0.4444 - accuracy: 0.9261 - mean_iou: 0.4849
Epoch 00001: accuracy did not improve from 0.94644
12/78 [===>..........................] - ETA: 50:36 - loss: 0.4463 - accuracy: 0.9251 - mean_iou: 0.4706
Epoch 00001: accuracy did not improve from 0.94644
13/78 [====>.........................] - ETA: 50:12 - loss: 0.4492 - accuracy: 0.9233 - mean_iou: 0.4605
Epoch 00001: accuracy did not improve from 0.94644
14/78 [====>.........................] - ETA: 49:44 - loss: 0.4476 - accuracy: 0.9236 - mean_iou: 0.4656
Epoch 00001: accuracy did not improve from 0.94644
15/78 [====>.........................] - ETA: 49:11 - loss: 0.4466 - accuracy: 0.9236 - mean_iou: 0.4614
Epoch 00001: accuracy did not improve from 0.94644
16/78 [=====>........................] - ETA: 48:38 - loss: 0.4407 - accuracy: 0.9250 - mean_iou: 0.4683
Epoch 00001: accuracy did not improve from 0.94644
17/78 [=====>........................] - ETA: 48:01 - loss: 0.4380 - accuracy: 0.9258 - mean_iou: 0.4768
Epoch 00001: accuracy did not improve from 0.94644
18/78 [=====>........................] - ETA: 47:24 - loss: 0.4350 - accuracy: 0.9261 - mean_iou: 0.4837
Epoch 00001: accuracy did not improve from 0.94644
19/78 [======>.......................] - ETA: 46:45 - loss: 0.4353 - accuracy: 0.9260 - mean_iou: 0.4837
Epoch 00001: accuracy did not improve from 0.94644
20/78 [======>.......................] - ETA: 46:08 - loss: 0.4316 - accuracy: 0.9268 - mean_iou: 0.4847
Epoch 00001: accuracy did not improve from 0.94644
21/78 [=======>......................] - ETA: 45:27 - loss: 0.4305 - accuracy: 0.9271 - mean_iou: 0.4883
Epoch 00001: accuracy did not improve from 0.94644
22/78 [=======>......................] - ETA: 44:45 - loss: 0.4291 - accuracy: 0.9276 - mean_iou: 0.4910
Epoch 00001: accuracy did not improve from 0.94644
23/78 [=======>......................] - ETA: 44:02 - loss: 0.4274 - accuracy: 0.9281 - mean_iou: 0.4962
Epoch 00001: accuracy did not improve from 0.94644
24/78 [========>.....................] - ETA: 43:19 - loss: 0.4263 - accuracy: 0.9282 - mean_iou: 0.4994
Epoch 00001: accuracy did not improve from 0.94644
25/78 [========>.....................] - ETA: 42:36 - loss: 0.4289 - accuracy: 0.9275 - mean_iou: 0.5004
Epoch 00001: accuracy did not improve from 0.94644
26/78 [=========>....................] - ETA: 41:51 - loss: 0.4273 - accuracy: 0.9277 - mean_iou: 0.5051
Epoch 00001: accuracy did not improve from 0.94644
27/78 [=========>....................] - ETA: 41:07 - loss: 0.4274 - accuracy: 0.9279 - mean_iou: 0.5091
Epoch 00001: accuracy did not improve from 0.94644
28/78 [=========>....................] - ETA: 40:22 - loss: 0.4284 - accuracy: 0.9274 - mean_iou: 0.5109
Epoch 00001: accuracy did not improve from 0.94644
29/78 [==========>...................] - ETA: 39:36 - loss: 0.4276 - accuracy: 0.9276 - mean_iou: 0.5112
Epoch 00001: accuracy did not improve from 0.94644
30/78 [==========>...................] - ETA: 38:51 - loss: 0.4268 - accuracy: 0.9278 - mean_iou: 0.5115
Epoch 00001: accuracy did not improve from 0.94644
31/78 [==========>...................] - ETA: 38:05 - loss: 0.4267 - accuracy: 0.9280 - mean_iou: 0.5137
Epoch 00001: accuracy did not improve from 0.94644
32/78 [===========>..................] - ETA: 37:18 - loss: 0.4263 - accuracy: 0.9283 - mean_iou: 0.5135
Epoch 00001: accuracy did not improve from 0.94644
33/78 [===========>..................] - ETA: 36:32 - loss: 0.4260 - accuracy: 0.9285 - mean_iou: 0.5133
Epoch 00001: accuracy did not improve from 0.94644
34/78 [============>.................] - ETA: 35:46 - loss: 0.4256 - accuracy: 0.9284 - mean_iou: 0.5128
Epoch 00001: accuracy did not improve from 0.94644
35/78 [============>.................] - ETA: 35:00 - loss: 0.4264 - accuracy: 0.9282 - mean_iou: 0.5133
Epoch 00001: accuracy did not improve from 0.94644
36/78 [============>.................] - ETA: 34:14 - loss: 0.4255 - accuracy: 0.9283 - mean_iou: 0.5156
Epoch 00001: accuracy did not improve from 0.94644
37/78 [=============>................] - ETA: 33:28 - loss: 0.4244 - accuracy: 0.9286 - mean_iou: 0.5140
Epoch 00001: accuracy did not improve from 0.94644
38/78 [=============>................] - ETA: 32:41 - loss: 0.4250 - accuracy: 0.9284 - mean_iou: 0.5150
Epoch 00001: accuracy did not improve from 0.94644
39/78 [==============>...............] - ETA: 31:53 - loss: 0.4254 - accuracy: 0.9280 - mean_iou: 0.5169
Epoch 00001: accuracy did not improve from 0.94644
40/78 [==============>...............] - ETA: 31:06 - loss: 0.4250 - accuracy: 0.9278 - mean_iou: 0.5168
Epoch 00001: accuracy did not improve from 0.94644
41/78 [==============>...............] - ETA: 30:18 - loss: 0.4244 - accuracy: 0.9280 - mean_iou: 0.5165
Epoch 00001: accuracy did not improve from 0.94644
42/78 [===============>..............] - ETA: 29:30 - loss: 0.4246 - accuracy: 0.9278 - mean_iou: 0.5127
Epoch 00001: accuracy did not improve from 0.94644
43/78 [===============>..............] - ETA: 28:42 - loss: 0.4247 - accuracy: 0.9276 - mean_iou: 0.5119
Epoch 00001: accuracy did not improve from 0.94644
44/78 [===============>..............] - ETA: 27:54 - loss: 0.4243 - accuracy: 0.9275 - mean_iou: 0.5115
Epoch 00001: accuracy did not improve from 0.94644
45/78 [================>.............] - ETA: 27:05 - loss: 0.4242 - accuracy: 0.9277 - mean_iou: 0.5125
Epoch 00001: accuracy did not improve from 0.94644
46/78 [================>.............] - ETA: 26:17 - loss: 0.4240 - accuracy: 0.9277 - mean_iou: 0.5131
Epoch 00001: accuracy did not improve from 0.94644
47/78 [=================>............] - ETA: 25:29 - loss: 0.4242 - accuracy: 0.9280 - mean_iou: 0.5129
Epoch 00001: accuracy did not improve from 0.94644
48/78 [=================>............] - ETA: 24:41 - loss: 0.4238 - accuracy: 0.9283 - mean_iou: 0.5163
Epoch 00001: accuracy did not improve from 0.94644
49/78 [=================>............] - ETA: 23:52 - loss: 0.4241 - accuracy: 0.9284 - mean_iou: 0.5180
Epoch 00001: accuracy did not improve from 0.94644
50/78 [==================>...........] - ETA: 23:03 - loss: 0.4246 - accuracy: 0.9284 - mean_iou: 0.5166
Epoch 00001: accuracy did not improve from 0.94644
51/78 [==================>...........] - ETA: 22:14 - loss: 0.4238 - accuracy: 0.9287 - mean_iou: 0.5197
Epoch 00001: accuracy did not improve from 0.94644
52/78 [===================>..........] - ETA: 21:25 - loss: 0.4238 - accuracy: 0.9289 - mean_iou: 0.5209
Epoch 00001: accuracy did not improve from 0.94644
53/78 [===================>..........] - ETA: 20:36 - loss: 0.4242 - accuracy: 0.9288 - mean_iou: 0.5206
Epoch 00001: accuracy did not improve from 0.94644
54/78 [===================>..........] - ETA: 19:47 - loss: 0.4243 - accuracy: 0.9287 - mean_iou: 0.5220
Epoch 00001: accuracy did not improve from 0.94644
55/78 [====================>.........] - ETA: 18:58 - loss: 0.4241 - accuracy: 0.9286 - mean_iou: 0.5228
Epoch 00001: accuracy did not improve from 0.94644
56/78 [====================>.........] - ETA: 18:09 - loss: 0.4240 - accuracy: 0.9287 - mean_iou: 0.5215
Epoch 00001: accuracy did not improve from 0.94644
57/78 [====================>.........] - ETA: 17:20 - loss: 0.4239 - accuracy: 0.9285 - mean_iou: 0.5206
Epoch 00001: accuracy did not improve from 0.94644
58/78 [=====================>........] - ETA: 16:31 - loss: 0.4242 - accuracy: 0.9284 - mean_iou: 0.5203
Epoch 00001: accuracy did not improve from 0.94644
59/78 [=====================>........] - ETA: 15:42 - loss: 0.4238 - accuracy: 0.9285 - mean_iou: 0.5198
Epoch 00001: accuracy did not improve from 0.94644
60/78 [======================>.......] - ETA: 14:53 - loss: 0.4246 - accuracy: 0.9282 - mean_iou: 0.5182
Epoch 00001: accuracy did not improve from 0.94644
61/78 [======================>.......] - ETA: 14:04 - loss: 0.4247 - accuracy: 0.9282 - mean_iou: 0.5166
Epoch 00001: accuracy did not improve from 0.94644
62/78 [======================>.......] - ETA: 13:15 - loss: 0.4242 - accuracy: 0.9285 - mean_iou: 0.5166
Epoch 00001: accuracy did not improve from 0.94644
63/78 [=======================>......] - ETA: 12:25 - loss: 0.4228 - accuracy: 0.9290 - mean_iou: 0.5194
Epoch 00001: accuracy did not improve from 0.94644
64/78 [=======================>......] - ETA: 11:36 - loss: 0.4231 - accuracy: 0.9288 - mean_iou: 0.5199
Epoch 00001: accuracy did not improve from 0.94644
65/78 [========================>.....] - ETA: 10:46 - loss: 0.4236 - accuracy: 0.9291 - mean_iou: 0.5202
Epoch 00001: accuracy did not improve from 0.94644
66/78 [========================>.....] - ETA: 9:57 - loss: 0.4249 - accuracy: 0.9288 - mean_iou: 0.5191 
Epoch 00001: accuracy did not improve from 0.94644
67/78 [========================>.....] - ETA: 9:07 - loss: 0.4245 - accuracy: 0.9290 - mean_iou: 0.5193
Epoch 00001: accuracy did not improve from 0.94644
68/78 [=========================>....] - ETA: 8:17 - loss: 0.4241 - accuracy: 0.9291 - mean_iou: 0.5207
Epoch 00001: accuracy did not improve from 0.94644
69/78 [=========================>....] - ETA: 7:27 - loss: 0.4240 - accuracy: 0.9291 - mean_iou: 0.5205
Epoch 00001: accuracy did not improve from 0.94644
70/78 [=========================>....] - ETA: 6:37 - loss: 0.4243 - accuracy: 0.9289 - mean_iou: 0.5190
Epoch 00001: accuracy did not improve from 0.94644
71/78 [==========================>...] - ETA: 5:47 - loss: 0.4239 - accuracy: 0.9290 - mean_iou: 0.5187
Epoch 00001: accuracy did not improve from 0.94644
72/78 [==========================>...] - ETA: 4:57 - loss: 0.4230 - accuracy: 0.9291 - mean_iou: 0.5198
Epoch 00001: accuracy did not improve from 0.94644
73/78 [===========================>..] - ETA: 4:07 - loss: 0.4230 - accuracy: 0.9291 - mean_iou: 0.5199
Epoch 00001: accuracy did not improve from 0.94644
74/78 [===========================>..] - ETA: 3:18 - loss: 0.4233 - accuracy: 0.9291 - mean_iou: 0.5202
Epoch 00001: accuracy did not improve from 0.94644
75/78 [===========================>..] - ETA: 2:28 - loss: 0.4237 - accuracy: 0.9287 - mean_iou: 0.5202
Epoch 00001: accuracy did not improve from 0.94644
76/78 [============================>.] - ETA: 1:38 - loss: 0.4241 - accuracy: 0.9287 - mean_iou: 0.5207
Epoch 00001: accuracy did not improve from 0.94644
77/78 [============================>.] - ETA: 49s - loss: 0.4241 - accuracy: 0.9288 - mean_iou: 0.5220 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.

Epoch 00001: accuracy did not improve from 0.94644
78/78 [==============================] - ETA: 0s - loss: 0.4237 - accuracy: 0.9288 - mean_iou: 0.5234 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
78/78 [==============================] - 4143s 53s/step - loss: 0.4237 - accuracy: 0.9288 - mean_iou: 0.5234 - val_loss: 0.4347 - val_accuracy: 0.9392 - val_mean_iou: 0.5553 - lr: 0.0010
Epoch 2/6
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.

Epoch 00002: accuracy did not improve from 0.94644
 1/78 [..............................] - ETA: 0s - loss: 0.3809 - accuracy: 0.9407 - mean_iou: 0.6340
Epoch 00002: accuracy did not improve from 0.94644
 2/78 [..............................] - ETA: 31:52 - loss: 0.4111 - accuracy: 0.9327 - mean_iou: 0.5802
Epoch 00002: accuracy did not improve from 0.94644
 3/78 [>.............................] - ETA: 41:58 - loss: 0.4265 - accuracy: 0.9267 - mean_iou: 0.5951
Epoch 00002: accuracy did not improve from 0.94644
 4/78 [>.............................] - ETA: 46:32 - loss: 0.4134 - accuracy: 0.9305 - mean_iou: 0.5844
Epoch 00002: accuracy did not improve from 0.94644
 5/78 [>.............................] - ETA: 49:05 - loss: 0.4139 - accuracy: 0.9314 - mean_iou: 0.6041
Epoch 00002: accuracy did not improve from 0.94644
 6/78 [=>............................] - ETA: 50:22 - loss: 0.4165 - accuracy: 0.9314 - mean_iou: 0.5777
Epoch 00002: accuracy did not improve from 0.94644
 7/78 [=>............................] - ETA: 51:05 - loss: 0.4193 - accuracy: 0.9297 - mean_iou: 0.5643
Epoch 00002: accuracy did not improve from 0.94644
 8/78 [==>...........................] - ETA: 51:25 - loss: 0.4182 - accuracy: 0.9286 - mean_iou: 0.5591
Epoch 00002: accuracy did not improve from 0.94644
 9/78 [==>...........................] - ETA: 51:32 - loss: 0.4182 - accuracy: 0.9289 - mean_iou: 0.5525
Epoch 00002: accuracy did not improve from 0.94644
10/78 [==>...........................] - ETA: 51:24 - loss: 0.4201 - accuracy: 0.9298 - mean_iou: 0.5448
Epoch 00002: accuracy did not improve from 0.94644
11/78 [===>..........................] - ETA: 51:14 - loss: 0.4205 - accuracy: 0.9306 - mean_iou: 0.5427
Epoch 00002: accuracy did not improve from 0.94644
12/78 [===>..........................] - ETA: 50:54 - loss: 0.4177 - accuracy: 0.9320 - mean_iou: 0.5411
Epoch 00002: accuracy did not improve from 0.94644
13/78 [====>.........................] - ETA: 50:29 - loss: 0.4163 - accuracy: 0.9326 - mean_iou: 0.5423
Epoch 00002: accuracy did not improve from 0.94644
14/78 [====>.........................] - ETA: 50:00 - loss: 0.4173 - accuracy: 0.9343 - mean_iou: 0.5503
Epoch 00002: accuracy did not improve from 0.94644
15/78 [====>.........................] - ETA: 49:27 - loss: 0.4157 - accuracy: 0.9342 - mean_iou: 0.5523
Epoch 00002: accuracy did not improve from 0.94644
16/78 [=====>........................] - ETA: 48:53 - loss: 0.4168 - accuracy: 0.9348 - mean_iou: 0.5532
Epoch 00002: accuracy did not improve from 0.94644
17/78 [=====>........................] - ETA: 48:18 - loss: 0.4138 - accuracy: 0.9355 - mean_iou: 0.5615
Epoch 00002: accuracy did not improve from 0.94644
18/78 [=====>........................] - ETA: 47:40 - loss: 0.4143 - accuracy: 0.9361 - mean_iou: 0.5638
Epoch 00002: accuracy did not improve from 0.94644
19/78 [======>.......................] - ETA: 47:00 - loss: 0.4120 - accuracy: 0.9368 - mean_iou: 0.5631
Epoch 00002: accuracy did not improve from 0.94644
20/78 [======>.......................] - ETA: 46:18 - loss: 0.4113 - accuracy: 0.9366 - mean_iou: 0.5634
Epoch 00002: accuracy did not improve from 0.94644
21/78 [=======>......................] - ETA: 45:38 - loss: 0.4087 - accuracy: 0.9368 - mean_iou: 0.5658
Epoch 00002: accuracy did not improve from 0.94644
22/78 [=======>......................] - ETA: 44:55 - loss: 0.4085 - accuracy: 0.9364 - mean_iou: 0.5722
Epoch 00002: accuracy did not improve from 0.94644
23/78 [=======>......................] - ETA: 44:13 - loss: 0.4098 - accuracy: 0.9356 - mean_iou: 0.5736
Epoch 00002: accuracy did not improve from 0.94644
24/78 [========>.....................] - ETA: 43:31 - loss: 0.4112 - accuracy: 0.9349 - mean_iou: 0.5665
Epoch 00002: accuracy did not improve from 0.94644
25/78 [========>.....................] - ETA: 42:47 - loss: 0.4117 - accuracy: 0.9348 - mean_iou: 0.5642
Epoch 00002: accuracy did not improve from 0.94644
26/78 [=========>....................] - ETA: 42:02 - loss: 0.4126 - accuracy: 0.9344 - mean_iou: 0.5625
Epoch 00002: accuracy did not improve from 0.94644
27/78 [=========>....................] - ETA: 41:17 - loss: 0.4140 - accuracy: 0.9334 - mean_iou: 0.5589
Epoch 00002: accuracy did not improve from 0.94644
28/78 [=========>....................] - ETA: 40:31 - loss: 0.4159 - accuracy: 0.9330 - mean_iou: 0.5551
Epoch 00002: accuracy did not improve from 0.94644
29/78 [==========>...................] - ETA: 39:46 - loss: 0.4163 - accuracy: 0.9329 - mean_iou: 0.5586
Epoch 00002: accuracy did not improve from 0.94644
30/78 [==========>...................] - ETA: 38:59 - loss: 0.4161 - accuracy: 0.9329 - mean_iou: 0.5583
Epoch 00002: accuracy did not improve from 0.94644
31/78 [==========>...................] - ETA: 38:12 - loss: 0.4156 - accuracy: 0.9333 - mean_iou: 0.5585
Epoch 00002: accuracy did not improve from 0.94644
32/78 [===========>..................] - ETA: 37:25 - loss: 0.4143 - accuracy: 0.9336 - mean_iou: 0.5611
Epoch 00002: accuracy did not improve from 0.94644
33/78 [===========>..................] - ETA: 36:37 - loss: 0.4148 - accuracy: 0.9335 - mean_iou: 0.5598
Epoch 00002: accuracy did not improve from 0.94644
34/78 [============>.................] - ETA: 35:49 - loss: 0.4145 - accuracy: 0.9340 - mean_iou: 0.5615
Epoch 00002: accuracy did not improve from 0.94644
35/78 [============>.................] - ETA: 35:02 - loss: 0.4149 - accuracy: 0.9337 - mean_iou: 0.5615
Epoch 00002: accuracy did not improve from 0.94644
36/78 [============>.................] - ETA: 34:15 - loss: 0.4140 - accuracy: 0.9338 - mean_iou: 0.5617
Epoch 00002: accuracy did not improve from 0.94644
37/78 [=============>................] - ETA: 33:27 - loss: 0.4128 - accuracy: 0.9343 - mean_iou: 0.5644
Epoch 00002: accuracy did not improve from 0.94644
38/78 [=============>................] - ETA: 32:39 - loss: 0.4141 - accuracy: 0.9339 - mean_iou: 0.5620
Epoch 00002: accuracy did not improve from 0.94644
39/78 [==============>...............] - ETA: 31:51 - loss: 0.4142 - accuracy: 0.9338 - mean_iou: 0.5598
Epoch 00002: accuracy did not improve from 0.94644
40/78 [==============>...............] - ETA: 31:03 - loss: 0.4141 - accuracy: 0.9337 - mean_iou: 0.5605
Epoch 00002: accuracy did not improve from 0.94644
41/78 [==============>...............] - ETA: 30:15 - loss: 0.4133 - accuracy: 0.9338 - mean_iou: 0.5604
Epoch 00002: accuracy did not improve from 0.94644
42/78 [===============>..............] - ETA: 29:27 - loss: 0.4130 - accuracy: 0.9337 - mean_iou: 0.5591
Epoch 00002: accuracy did not improve from 0.94644
43/78 [===============>..............] - ETA: 28:39 - loss: 0.4119 - accuracy: 0.9336 - mean_iou: 0.5624
Epoch 00002: accuracy did not improve from 0.94644
44/78 [===============>..............] - ETA: 27:50 - loss: 0.4107 - accuracy: 0.9337 - mean_iou: 0.5636
Epoch 00002: accuracy did not improve from 0.94644
45/78 [================>.............] - ETA: 27:02 - loss: 0.4103 - accuracy: 0.9337 - mean_iou: 0.5650
Epoch 00002: accuracy did not improve from 0.94644
46/78 [================>.............] - ETA: 26:13 - loss: 0.4110 - accuracy: 0.9335 - mean_iou: 0.5632
Epoch 00002: accuracy did not improve from 0.94644
47/78 [=================>............] - ETA: 25:24 - loss: 0.4107 - accuracy: 0.9334 - mean_iou: 0.5623
Epoch 00002: accuracy did not improve from 0.94644
48/78 [=================>............] - ETA: 24:36 - loss: 0.4113 - accuracy: 0.9334 - mean_iou: 0.5620
Epoch 00002: accuracy did not improve from 0.94644
49/78 [=================>............] - ETA: 23:47 - loss: 0.4104 - accuracy: 0.9334 - mean_iou: 0.5634
Epoch 00002: accuracy did not improve from 0.94644
50/78 [==================>...........] - ETA: 22:59 - loss: 0.4100 - accuracy: 0.9334 - mean_iou: 0.5647
Epoch 00002: accuracy did not improve from 0.94644
51/78 [==================>...........] - ETA: 22:10 - loss: 0.4108 - accuracy: 0.9329 - mean_iou: 0.5627
Epoch 00002: accuracy did not improve from 0.94644
52/78 [===================>..........] - ETA: 21:21 - loss: 0.4107 - accuracy: 0.9328 - mean_iou: 0.5617
Epoch 00002: accuracy did not improve from 0.94644
53/78 [===================>..........] - ETA: 20:32 - loss: 0.4103 - accuracy: 0.9327 - mean_iou: 0.5631
Epoch 00002: accuracy did not improve from 0.94644
54/78 [===================>..........] - ETA: 19:43 - loss: 0.4105 - accuracy: 0.9324 - mean_iou: 0.5624
Epoch 00002: accuracy did not improve from 0.94644
55/78 [====================>.........] - ETA: 18:54 - loss: 0.4114 - accuracy: 0.9322 - mean_iou: 0.5612
Epoch 00002: accuracy did not improve from 0.94644
56/78 [====================>.........] - ETA: 18:05 - loss: 0.4114 - accuracy: 0.9319 - mean_iou: 0.5592
Epoch 00002: accuracy did not improve from 0.94644
57/78 [====================>.........] - ETA: 17:16 - loss: 0.4116 - accuracy: 0.9316 - mean_iou: 0.5572
Epoch 00002: accuracy did not improve from 0.94644
58/78 [=====================>........] - ETA: 16:27 - loss: 0.4121 - accuracy: 0.9312 - mean_iou: 0.5556
Epoch 00002: accuracy did not improve from 0.94644
59/78 [=====================>........] - ETA: 15:38 - loss: 0.4113 - accuracy: 0.9314 - mean_iou: 0.5543
Epoch 00002: accuracy did not improve from 0.94644
60/78 [======================>.......] - ETA: 14:49 - loss: 0.4110 - accuracy: 0.9315 - mean_iou: 0.5531
Epoch 00002: accuracy did not improve from 0.94644
61/78 [======================>.......] - ETA: 14:00 - loss: 0.4105 - accuracy: 0.9316 - mean_iou: 0.5544
Epoch 00002: accuracy did not improve from 0.94644
62/78 [======================>.......] - ETA: 13:11 - loss: 0.4102 - accuracy: 0.9317 - mean_iou: 0.5538
Epoch 00002: accuracy did not improve from 0.94644
63/78 [=======================>......] - ETA: 12:21 - loss: 0.4104 - accuracy: 0.9316 - mean_iou: 0.5542
Epoch 00002: accuracy did not improve from 0.94644
64/78 [=======================>......] - ETA: 11:32 - loss: 0.4093 - accuracy: 0.9319 - mean_iou: 0.5551
Epoch 00002: accuracy did not improve from 0.94644
65/78 [========================>.....] - ETA: 10:43 - loss: 0.4098 - accuracy: 0.9319 - mean_iou: 0.5533
Epoch 00002: accuracy did not improve from 0.94644
66/78 [========================>.....] - ETA: 9:53 - loss: 0.4094 - accuracy: 0.9320 - mean_iou: 0.5546 
Epoch 00002: accuracy did not improve from 0.94644
67/78 [========================>.....] - ETA: 9:04 - loss: 0.4101 - accuracy: 0.9318 - mean_iou: 0.5545
Epoch 00002: accuracy did not improve from 0.94644
68/78 [=========================>....] - ETA: 8:14 - loss: 0.4099 - accuracy: 0.9317 - mean_iou: 0.5542
Epoch 00002: accuracy did not improve from 0.94644
69/78 [=========================>....] - ETA: 7:24 - loss: 0.4101 - accuracy: 0.9318 - mean_iou: 0.5528
Epoch 00002: accuracy did not improve from 0.94644
70/78 [=========================>....] - ETA: 6:34 - loss: 0.4102 - accuracy: 0.9317 - mean_iou: 0.5511
Epoch 00002: accuracy did not improve from 0.94644
71/78 [==========================>...] - ETA: 5:45 - loss: 0.4105 - accuracy: 0.9317 - mean_iou: 0.5506
Epoch 00002: accuracy did not improve from 0.94644
72/78 [==========================>...] - ETA: 4:55 - loss: 0.4107 - accuracy: 0.9318 - mean_iou: 0.5511
Epoch 00002: accuracy did not improve from 0.94644
73/78 [===========================>..] - ETA: 4:06 - loss: 0.4114 - accuracy: 0.9318 - mean_iou: 0.5493
Epoch 00002: accuracy did not improve from 0.94644
74/78 [===========================>..] - ETA: 3:16 - loss: 0.4113 - accuracy: 0.9319 - mean_iou: 0.5514
Epoch 00002: accuracy did not improve from 0.94644
75/78 [===========================>..] - ETA: 2:27 - loss: 0.4110 - accuracy: 0.9321 - mean_iou: 0.5522
Epoch 00002: accuracy did not improve from 0.94644
76/78 [============================>.] - ETA: 1:38 - loss: 0.4114 - accuracy: 0.9320 - mean_iou: 0.5508
Epoch 00002: accuracy did not improve from 0.94644
77/78 [============================>.] - ETA: 49s - loss: 0.4110 - accuracy: 0.9323 - mean_iou: 0.5515 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.

Epoch 00002: accuracy did not improve from 0.94644
78/78 [==============================] - ETA: 0s - loss: 0.4117 - accuracy: 0.9320 - mean_iou: 0.5512 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
78/78 [==============================] - 4111s 53s/step - loss: 0.4117 - accuracy: 0.9320 - mean_iou: 0.5512 - val_loss: 0.4324 - val_accuracy: 0.9253 - val_mean_iou: 0.4669 - lr: 9.9606e-04
Epoch 3/6
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.

Epoch 00003: accuracy did not improve from 0.94644
 1/78 [..............................] - ETA: 0s - loss: 0.3943 - accuracy: 0.9429 - mean_iou: 0.5751
Epoch 00003: accuracy did not improve from 0.94644
 2/78 [..............................] - ETA: 31:48 - loss: 0.4072 - accuracy: 0.9425 - mean_iou: 0.5413
Epoch 00003: accuracy improved from 0.94644 to 0.94873, saving model to 4000_checkpoint_1.hdf5
 3/78 [>.............................] - ETA: 42:41 - loss: 0.3889 - accuracy: 0.9487 - mean_iou: 0.5784
Epoch 00003: accuracy did not improve from 0.94873
 4/78 [>.............................] - ETA: 47:12 - loss: 0.3966 - accuracy: 0.9465 - mean_iou: 0.5681
Epoch 00003: accuracy did not improve from 0.94873
 5/78 [>.............................] - ETA: 49:27 - loss: 0.4002 - accuracy: 0.9446 - mean_iou: 0.5757
Epoch 00003: accuracy did not improve from 0.94873
 6/78 [=>............................] - ETA: 50:39 - loss: 0.4052 - accuracy: 0.9460 - mean_iou: 0.5821
Epoch 00003: accuracy did not improve from 0.94873
 7/78 [=>............................] - ETA: 51:15 - loss: 0.3991 - accuracy: 0.9462 - mean_iou: 0.5858
Epoch 00003: accuracy did not improve from 0.94873
 8/78 [==>...........................] - ETA: 51:31 - loss: 0.3915 - accuracy: 0.9473 - mean_iou: 0.5896
Epoch 00003: accuracy did not improve from 0.94873
 9/78 [==>...........................] - ETA: 51:41 - loss: 0.3899 - accuracy: 0.9468 - mean_iou: 0.5961
Epoch 00003: accuracy did not improve from 0.94873
10/78 [==>...........................] - ETA: 51:34 - loss: 0.3907 - accuracy: 0.9453 - mean_iou: 0.5937
Epoch 00003: accuracy did not improve from 0.94873
11/78 [===>..........................] - ETA: 51:19 - loss: 0.3919 - accuracy: 0.9435 - mean_iou: 0.5924
Epoch 00003: accuracy did not improve from 0.94873
12/78 [===>..........................] - ETA: 51:01 - loss: 0.3922 - accuracy: 0.9419 - mean_iou: 0.5909
Epoch 00003: accuracy did not improve from 0.94873
13/78 [====>.........................] - ETA: 50:37 - loss: 0.3920 - accuracy: 0.9410 - mean_iou: 0.5877
Epoch 00003: accuracy did not improve from 0.94873
14/78 [====>.........................] - ETA: 50:07 - loss: 0.3950 - accuracy: 0.9391 - mean_iou: 0.5780
Epoch 00003: accuracy did not improve from 0.94873
15/78 [====>.........................] - ETA: 49:34 - loss: 0.3960 - accuracy: 0.9379 - mean_iou: 0.5716
Epoch 00003: accuracy did not improve from 0.94873
16/78 [=====>........................] - ETA: 49:02 - loss: 0.3966 - accuracy: 0.9367 - mean_iou: 0.5733
Epoch 00003: accuracy did not improve from 0.94873
17/78 [=====>........................] - ETA: 48:26 - loss: 0.3954 - accuracy: 0.9359 - mean_iou: 0.5700
Epoch 00003: accuracy did not improve from 0.94873
18/78 [=====>........................] - ETA: 47:48 - loss: 0.3941 - accuracy: 0.9355 - mean_iou: 0.5703
Epoch 00003: accuracy did not improve from 0.94873
19/78 [======>.......................] - ETA: 47:09 - loss: 0.3917 - accuracy: 0.9356 - mean_iou: 0.5742
Epoch 00003: accuracy did not improve from 0.94873
20/78 [======>.......................] - ETA: 46:29 - loss: 0.3906 - accuracy: 0.9356 - mean_iou: 0.5780
Epoch 00003: accuracy did not improve from 0.94873
21/78 [=======>......................] - ETA: 45:47 - loss: 0.3899 - accuracy: 0.9356 - mean_iou: 0.5789
Epoch 00003: accuracy did not improve from 0.94873
22/78 [=======>......................] - ETA: 45:05 - loss: 0.3886 - accuracy: 0.9354 - mean_iou: 0.5850
Epoch 00003: accuracy did not improve from 0.94873
23/78 [=======>......................] - ETA: 44:22 - loss: 0.3879 - accuracy: 0.9355 - mean_iou: 0.5848
Epoch 00003: accuracy did not improve from 0.94873
24/78 [========>.....................] - ETA: 43:38 - loss: 0.3882 - accuracy: 0.9358 - mean_iou: 0.5859
Epoch 00003: accuracy did not improve from 0.94873
25/78 [========>.....................] - ETA: 42:55 - loss: 0.3876 - accuracy: 0.9358 - mean_iou: 0.5844
Epoch 00003: accuracy did not improve from 0.94873
26/78 [=========>....................] - ETA: 42:10 - loss: 0.3885 - accuracy: 0.9357 - mean_iou: 0.5839
Epoch 00003: accuracy did not improve from 0.94873
27/78 [=========>....................] - ETA: 41:25 - loss: 0.3886 - accuracy: 0.9355 - mean_iou: 0.5811
Epoch 00003: accuracy did not improve from 0.94873
28/78 [=========>....................] - ETA: 40:40 - loss: 0.3887 - accuracy: 0.9358 - mean_iou: 0.5812
Epoch 00003: accuracy did not improve from 0.94873
29/78 [==========>...................] - ETA: 39:54 - loss: 0.3901 - accuracy: 0.9361 - mean_iou: 0.5797
Epoch 00003: accuracy did not improve from 0.94873
30/78 [==========>...................] - ETA: 39:08 - loss: 0.3896 - accuracy: 0.9364 - mean_iou: 0.5797
Epoch 00003: accuracy did not improve from 0.94873
31/78 [==========>...................] - ETA: 38:22 - loss: 0.3887 - accuracy: 0.9367 - mean_iou: 0.5808
Epoch 00003: accuracy did not improve from 0.94873
32/78 [===========>..................] - ETA: 37:35 - loss: 0.3888 - accuracy: 0.9366 - mean_iou: 0.5832
Epoch 00003: accuracy did not improve from 0.94873
33/78 [===========>..................] - ETA: 36:48 - loss: 0.3881 - accuracy: 0.9367 - mean_iou: 0.5846
Epoch 00003: accuracy did not improve from 0.94873
34/78 [============>.................] - ETA: 36:02 - loss: 0.3873 - accuracy: 0.9371 - mean_iou: 0.5857
Epoch 00003: accuracy did not improve from 0.94873
35/78 [============>.................] - ETA: 35:15 - loss: 0.3890 - accuracy: 0.9373 - mean_iou: 0.5851
Epoch 00003: accuracy did not improve from 0.94873
36/78 [============>.................] - ETA: 34:27 - loss: 0.3876 - accuracy: 0.9376 - mean_iou: 0.5875
Epoch 00003: accuracy did not improve from 0.94873
37/78 [=============>................] - ETA: 33:40 - loss: 0.3881 - accuracy: 0.9374 - mean_iou: 0.5871
Epoch 00003: accuracy did not improve from 0.94873
38/78 [=============>................] - ETA: 32:53 - loss: 0.3885 - accuracy: 0.9373 - mean_iou: 0.5857
Epoch 00003: accuracy did not improve from 0.94873
39/78 [==============>...............] - ETA: 32:05 - loss: 0.3898 - accuracy: 0.9371 - mean_iou: 0.5843
Epoch 00003: accuracy did not improve from 0.94873
40/78 [==============>...............] - ETA: 31:16 - loss: 0.3895 - accuracy: 0.9372 - mean_iou: 0.5837
Epoch 00003: accuracy did not improve from 0.94873
41/78 [==============>...............] - ETA: 30:28 - loss: 0.3891 - accuracy: 0.9372 - mean_iou: 0.5849
Epoch 00003: accuracy did not improve from 0.94873
42/78 [===============>..............] - ETA: 29:40 - loss: 0.3905 - accuracy: 0.9368 - mean_iou: 0.5834
Epoch 00003: accuracy did not improve from 0.94873
43/78 [===============>..............] - ETA: 28:51 - loss: 0.3902 - accuracy: 0.9367 - mean_iou: 0.5825
Epoch 00003: accuracy did not improve from 0.94873
44/78 [===============>..............] - ETA: 28:02 - loss: 0.3902 - accuracy: 0.9368 - mean_iou: 0.5845
Epoch 00003: accuracy did not improve from 0.94873
45/78 [================>.............] - ETA: 27:13 - loss: 0.3894 - accuracy: 0.9369 - mean_iou: 0.5829
Epoch 00003: accuracy did not improve from 0.94873
46/78 [================>.............] - ETA: 26:24 - loss: 0.3890 - accuracy: 0.9370 - mean_iou: 0.5815
Epoch 00003: accuracy did not improve from 0.94873
47/78 [=================>............] - ETA: 25:36 - loss: 0.3894 - accuracy: 0.9367 - mean_iou: 0.5795
Epoch 00003: accuracy did not improve from 0.94873
48/78 [=================>............] - ETA: 24:47 - loss: 0.3896 - accuracy: 0.9367 - mean_iou: 0.5804
Epoch 00003: accuracy did not improve from 0.94873
49/78 [=================>............] - ETA: 23:58 - loss: 0.3890 - accuracy: 0.9369 - mean_iou: 0.5805
Epoch 00003: accuracy did not improve from 0.94873
50/78 [==================>...........] - ETA: 23:09 - loss: 0.3911 - accuracy: 0.9366 - mean_iou: 0.5771
Epoch 00003: accuracy did not improve from 0.94873
51/78 [==================>...........] - ETA: 22:20 - loss: 0.3907 - accuracy: 0.9365 - mean_iou: 0.5786
Epoch 00003: accuracy did not improve from 0.94873
52/78 [===================>..........] - ETA: 21:31 - loss: 0.3906 - accuracy: 0.9365 - mean_iou: 0.5798
Epoch 00003: accuracy did not improve from 0.94873
53/78 [===================>..........] - ETA: 20:41 - loss: 0.3898 - accuracy: 0.9367 - mean_iou: 0.5803
Epoch 00003: accuracy did not improve from 0.94873
54/78 [===================>..........] - ETA: 19:52 - loss: 0.3901 - accuracy: 0.9369 - mean_iou: 0.5819
Epoch 00003: accuracy did not improve from 0.94873
55/78 [====================>.........] - ETA: 19:03 - loss: 0.3892 - accuracy: 0.9371 - mean_iou: 0.5840
Epoch 00003: accuracy did not improve from 0.94873
56/78 [====================>.........] - ETA: 18:13 - loss: 0.3892 - accuracy: 0.9369 - mean_iou: 0.5840
Epoch 00003: accuracy did not improve from 0.94873
57/78 [====================>.........] - ETA: 17:24 - loss: 0.3885 - accuracy: 0.9370 - mean_iou: 0.5856
Epoch 00003: accuracy did not improve from 0.94873
58/78 [=====================>........] - ETA: 16:35 - loss: 0.3894 - accuracy: 0.9368 - mean_iou: 0.5847
Epoch 00003: accuracy did not improve from 0.94873
59/78 [=====================>........] - ETA: 15:45 - loss: 0.3885 - accuracy: 0.9370 - mean_iou: 0.5871
Epoch 00003: accuracy did not improve from 0.94873
60/78 [======================>.......] - ETA: 14:56 - loss: 0.3893 - accuracy: 0.9368 - mean_iou: 0.5857
Epoch 00003: accuracy did not improve from 0.94873
61/78 [======================>.......] - ETA: 14:06 - loss: 0.3893 - accuracy: 0.9367 - mean_iou: 0.5856
Epoch 00003: accuracy did not improve from 0.94873
62/78 [======================>.......] - ETA: 13:17 - loss: 0.3894 - accuracy: 0.9366 - mean_iou: 0.5847
Epoch 00003: accuracy did not improve from 0.94873
63/78 [=======================>......] - ETA: 12:27 - loss: 0.3896 - accuracy: 0.9366 - mean_iou: 0.5836
Epoch 00003: accuracy did not improve from 0.94873
64/78 [=======================>......] - ETA: 11:37 - loss: 0.3894 - accuracy: 0.9366 - mean_iou: 0.5836
Epoch 00003: accuracy did not improve from 0.94873
65/78 [========================>.....] - ETA: 10:48 - loss: 0.3891 - accuracy: 0.9366 - mean_iou: 0.5844
Epoch 00003: accuracy did not improve from 0.94873
66/78 [========================>.....] - ETA: 9:58 - loss: 0.3903 - accuracy: 0.9361 - mean_iou: 0.5834 
Epoch 00003: accuracy did not improve from 0.94873
67/78 [========================>.....] - ETA: 9:08 - loss: 0.3895 - accuracy: 0.9363 - mean_iou: 0.5842
Epoch 00003: accuracy did not improve from 0.94873
68/78 [=========================>....] - ETA: 8:18 - loss: 0.3891 - accuracy: 0.9364 - mean_iou: 0.5842
Epoch 00003: accuracy did not improve from 0.94873
69/78 [=========================>....] - ETA: 7:28 - loss: 0.3889 - accuracy: 0.9365 - mean_iou: 0.5835
Epoch 00003: accuracy did not improve from 0.94873
70/78 [=========================>....] - ETA: 6:37 - loss: 0.3900 - accuracy: 0.9364 - mean_iou: 0.5816
Epoch 00003: accuracy did not improve from 0.94873
71/78 [==========================>...] - ETA: 5:47 - loss: 0.3901 - accuracy: 0.9363 - mean_iou: 0.5806
Epoch 00003: accuracy did not improve from 0.94873
72/78 [==========================>...] - ETA: 4:57 - loss: 0.3901 - accuracy: 0.9364 - mean_iou: 0.5805
Epoch 00003: accuracy did not improve from 0.94873
73/78 [===========================>..] - ETA: 4:08 - loss: 0.3902 - accuracy: 0.9364 - mean_iou: 0.5795
Epoch 00003: accuracy did not improve from 0.94873
74/78 [===========================>..] - ETA: 3:18 - loss: 0.3901 - accuracy: 0.9363 - mean_iou: 0.5798
Epoch 00003: accuracy did not improve from 0.94873
75/78 [===========================>..] - ETA: 2:28 - loss: 0.3909 - accuracy: 0.9363 - mean_iou: 0.5795
Epoch 00003: accuracy did not improve from 0.94873
76/78 [============================>.] - ETA: 1:38 - loss: 0.3914 - accuracy: 0.9362 - mean_iou: 0.5798
Epoch 00003: accuracy did not improve from 0.94873
77/78 [============================>.] - ETA: 49s - loss: 0.3919 - accuracy: 0.9362 - mean_iou: 0.5786 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.

Epoch 00003: accuracy did not improve from 0.94873
78/78 [==============================] - ETA: 0s - loss: 0.3919 - accuracy: 0.9362 - mean_iou: 0.5779 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
78/78 [==============================] - 4141s 53s/step - loss: 0.3919 - accuracy: 0.9362 - mean_iou: 0.5779 - val_loss: 0.4271 - val_accuracy: 0.9390 - val_mean_iou: 0.5421 - lr: 9.8429e-04
Epoch 4/6
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.

Epoch 00004: accuracy did not improve from 0.94873
 1/78 [..............................] - ETA: 0s - loss: 0.3876 - accuracy: 0.9384 - mean_iou: 0.5804
Epoch 00004: accuracy did not improve from 0.94873
 2/78 [..............................] - ETA: 32:16 - loss: 0.4070 - accuracy: 0.9361 - mean_iou: 0.5507
Epoch 00004: accuracy did not improve from 0.94873
 3/78 [>.............................] - ETA: 42:25 - loss: 0.4017 - accuracy: 0.9353 - mean_iou: 0.5642
Epoch 00004: accuracy did not improve from 0.94873
 4/78 [>.............................] - ETA: 47:05 - loss: 0.3877 - accuracy: 0.9381 - mean_iou: 0.5860
Epoch 00004: accuracy did not improve from 0.94873
 5/78 [>.............................] - ETA: 49:40 - loss: 0.3888 - accuracy: 0.9358 - mean_iou: 0.5945
Epoch 00004: accuracy did not improve from 0.94873
 6/78 [=>............................] - ETA: 50:54 - loss: 0.3840 - accuracy: 0.9372 - mean_iou: 0.5990
Epoch 00004: accuracy did not improve from 0.94873
 7/78 [=>............................] - ETA: 51:36 - loss: 0.3993 - accuracy: 0.9328 - mean_iou: 0.5746
Epoch 00004: accuracy did not improve from 0.94873
 8/78 [==>...........................] - ETA: 51:52 - loss: 0.3973 - accuracy: 0.9330 - mean_iou: 0.5850
Epoch 00004: accuracy did not improve from 0.94873
 9/78 [==>...........................] - ETA: 51:54 - loss: 0.3929 - accuracy: 0.9331 - mean_iou: 0.5911
Epoch 00004: accuracy did not improve from 0.94873
10/78 [==>...........................] - ETA: 51:47 - loss: 0.3918 - accuracy: 0.9326 - mean_iou: 0.5955
Epoch 00004: accuracy did not improve from 0.94873
11/78 [===>..........................] - ETA: 51:34 - loss: 0.3933 - accuracy: 0.9323 - mean_iou: 0.5939
Epoch 00004: accuracy did not improve from 0.94873
12/78 [===>..........................] - ETA: 51:11 - loss: 0.3939 - accuracy: 0.9314 - mean_iou: 0.5933
Epoch 00004: accuracy did not improve from 0.94873
13/78 [====>.........................] - ETA: 50:44 - loss: 0.3919 - accuracy: 0.9311 - mean_iou: 0.5906
Epoch 00004: accuracy did not improve from 0.94873
14/78 [====>.........................] - ETA: 50:13 - loss: 0.3944 - accuracy: 0.9307 - mean_iou: 0.5876
Epoch 00004: accuracy did not improve from 0.94873
15/78 [====>.........................] - ETA: 49:39 - loss: 0.3997 - accuracy: 0.9295 - mean_iou: 0.5811
Epoch 00004: accuracy did not improve from 0.94873
16/78 [=====>........................] - ETA: 49:06 - loss: 0.4031 - accuracy: 0.9294 - mean_iou: 0.5731
Epoch 00004: accuracy did not improve from 0.94873
17/78 [=====>........................] - ETA: 48:29 - loss: 0.4010 - accuracy: 0.9300 - mean_iou: 0.5721
Epoch 00004: accuracy did not improve from 0.94873
18/78 [=====>........................] - ETA: 47:51 - loss: 0.4008 - accuracy: 0.9307 - mean_iou: 0.5653
Epoch 00004: accuracy did not improve from 0.94873
19/78 [======>.......................] - ETA: 47:12 - loss: 0.4017 - accuracy: 0.9314 - mean_iou: 0.5630
Epoch 00004: accuracy did not improve from 0.94873
20/78 [======>.......................] - ETA: 46:31 - loss: 0.4006 - accuracy: 0.9316 - mean_iou: 0.5653
Epoch 00004: accuracy did not improve from 0.94873
21/78 [=======>......................] - ETA: 45:49 - loss: 0.3994 - accuracy: 0.9324 - mean_iou: 0.5709
Epoch 00004: accuracy did not improve from 0.94873
22/78 [=======>......................] - ETA: 45:07 - loss: 0.3991 - accuracy: 0.9328 - mean_iou: 0.5714
Epoch 00004: accuracy did not improve from 0.94873
23/78 [=======>......................] - ETA: 44:25 - loss: 0.3968 - accuracy: 0.9339 - mean_iou: 0.5762
Epoch 00004: accuracy did not improve from 0.94873
24/78 [========>.....................] - ETA: 43:41 - loss: 0.3957 - accuracy: 0.9346 - mean_iou: 0.5796
Epoch 00004: accuracy did not improve from 0.94873
25/78 [========>.....................] - ETA: 42:57 - loss: 0.3972 - accuracy: 0.9342 - mean_iou: 0.5792
Epoch 00004: accuracy did not improve from 0.94873
26/78 [=========>....................] - ETA: 42:11 - loss: 0.3964 - accuracy: 0.9346 - mean_iou: 0.5775
Epoch 00004: accuracy did not improve from 0.94873
27/78 [=========>....................] - ETA: 41:25 - loss: 0.3944 - accuracy: 0.9351 - mean_iou: 0.5776
Epoch 00004: accuracy did not improve from 0.94873
28/78 [=========>....................] - ETA: 40:39 - loss: 0.3939 - accuracy: 0.9350 - mean_iou: 0.5783
Epoch 00004: accuracy did not improve from 0.94873
29/78 [==========>...................] - ETA: 39:53 - loss: 0.3944 - accuracy: 0.9346 - mean_iou: 0.5781
Epoch 00004: accuracy did not improve from 0.94873
30/78 [==========>...................] - ETA: 39:07 - loss: 0.3928 - accuracy: 0.9349 - mean_iou: 0.5792
Epoch 00004: accuracy did not improve from 0.94873
31/78 [==========>...................] - ETA: 38:20 - loss: 0.3931 - accuracy: 0.9346 - mean_iou: 0.5776
Epoch 00004: accuracy did not improve from 0.94873
32/78 [===========>..................] - ETA: 37:33 - loss: 0.3956 - accuracy: 0.9346 - mean_iou: 0.5761
Epoch 00004: accuracy did not improve from 0.94873
33/78 [===========>..................] - ETA: 36:46 - loss: 0.3950 - accuracy: 0.9343 - mean_iou: 0.5759
Epoch 00004: accuracy did not improve from 0.94873
34/78 [============>.................] - ETA: 35:58 - loss: 0.3936 - accuracy: 0.9345 - mean_iou: 0.5776
Epoch 00004: accuracy did not improve from 0.94873
35/78 [============>.................] - ETA: 35:11 - loss: 0.3951 - accuracy: 0.9344 - mean_iou: 0.5787
Epoch 00004: accuracy did not improve from 0.94873
36/78 [============>.................] - ETA: 34:24 - loss: 0.3947 - accuracy: 0.9346 - mean_iou: 0.5775
Epoch 00004: accuracy did not improve from 0.94873
37/78 [=============>................] - ETA: 33:35 - loss: 0.3929 - accuracy: 0.9350 - mean_iou: 0.5795
Epoch 00004: accuracy did not improve from 0.94873
38/78 [=============>................] - ETA: 32:48 - loss: 0.3928 - accuracy: 0.9352 - mean_iou: 0.5812
Epoch 00004: accuracy did not improve from 0.94873
39/78 [==============>...............] - ETA: 32:00 - loss: 0.3926 - accuracy: 0.9354 - mean_iou: 0.5793
Epoch 00004: accuracy did not improve from 0.94873
40/78 [==============>...............] - ETA: 31:12 - loss: 0.3919 - accuracy: 0.9355 - mean_iou: 0.5812
Epoch 00004: accuracy did not improve from 0.94873
41/78 [==============>...............] - ETA: 30:24 - loss: 0.3931 - accuracy: 0.9357 - mean_iou: 0.5809
Epoch 00004: accuracy did not improve from 0.94873
42/78 [===============>..............] - ETA: 29:36 - loss: 0.3930 - accuracy: 0.9356 - mean_iou: 0.5831
Epoch 00004: accuracy did not improve from 0.94873
43/78 [===============>..............] - ETA: 28:47 - loss: 0.3939 - accuracy: 0.9355 - mean_iou: 0.5823
Epoch 00004: accuracy did not improve from 0.94873
44/78 [===============>..............] - ETA: 27:59 - loss: 0.3933 - accuracy: 0.9358 - mean_iou: 0.5847
Epoch 00004: accuracy did not improve from 0.94873
45/78 [================>.............] - ETA: 27:10 - loss: 0.3929 - accuracy: 0.9361 - mean_iou: 0.5858
Epoch 00004: accuracy did not improve from 0.94873
46/78 [================>.............] - ETA: 26:22 - loss: 0.3919 - accuracy: 0.9362 - mean_iou: 0.5870
Epoch 00004: accuracy did not improve from 0.94873
47/78 [=================>............] - ETA: 25:33 - loss: 0.3908 - accuracy: 0.9365 - mean_iou: 0.5883
Epoch 00004: accuracy did not improve from 0.94873
48/78 [=================>............] - ETA: 24:45 - loss: 0.3907 - accuracy: 0.9365 - mean_iou: 0.5892
Epoch 00004: accuracy did not improve from 0.94873
49/78 [=================>............] - ETA: 23:56 - loss: 0.3914 - accuracy: 0.9365 - mean_iou: 0.5887
Epoch 00004: accuracy did not improve from 0.94873
50/78 [==================>...........] - ETA: 23:07 - loss: 0.3913 - accuracy: 0.9366 - mean_iou: 0.5880
Epoch 00004: accuracy did not improve from 0.94873
51/78 [==================>...........] - ETA: 22:18 - loss: 0.3909 - accuracy: 0.9368 - mean_iou: 0.5872
Epoch 00004: accuracy did not improve from 0.94873
52/78 [===================>..........] - ETA: 21:29 - loss: 0.3891 - accuracy: 0.9373 - mean_iou: 0.5898
Epoch 00004: accuracy did not improve from 0.94873
53/78 [===================>..........] - ETA: 20:40 - loss: 0.3888 - accuracy: 0.9375 - mean_iou: 0.5908
Epoch 00004: accuracy did not improve from 0.94873
54/78 [===================>..........] - ETA: 19:50 - loss: 0.3886 - accuracy: 0.9376 - mean_iou: 0.5900
Epoch 00004: accuracy did not improve from 0.94873
55/78 [====================>.........] - ETA: 19:01 - loss: 0.3894 - accuracy: 0.9374 - mean_iou: 0.5885
Epoch 00004: accuracy did not improve from 0.94873
56/78 [====================>.........] - ETA: 18:12 - loss: 0.3888 - accuracy: 0.9376 - mean_iou: 0.5898
Epoch 00004: accuracy did not improve from 0.94873
57/78 [====================>.........] - ETA: 17:22 - loss: 0.3890 - accuracy: 0.9377 - mean_iou: 0.5899
Epoch 00004: accuracy did not improve from 0.94873
58/78 [=====================>........] - ETA: 16:33 - loss: 0.3895 - accuracy: 0.9375 - mean_iou: 0.5894
Epoch 00004: accuracy did not improve from 0.94873
59/78 [=====================>........] - ETA: 15:44 - loss: 0.3901 - accuracy: 0.9373 - mean_iou: 0.5874
Epoch 00004: accuracy did not improve from 0.94873
60/78 [======================>.......] - ETA: 14:54 - loss: 0.3894 - accuracy: 0.9374 - mean_iou: 0.5871
Epoch 00004: accuracy did not improve from 0.94873
61/78 [======================>.......] - ETA: 14:05 - loss: 0.3894 - accuracy: 0.9373 - mean_iou: 0.5867
Epoch 00004: accuracy did not improve from 0.94873
62/78 [======================>.......] - ETA: 13:15 - loss: 0.3894 - accuracy: 0.9373 - mean_iou: 0.5846
Epoch 00004: accuracy did not improve from 0.94873
63/78 [=======================>......] - ETA: 12:26 - loss: 0.3899 - accuracy: 0.9371 - mean_iou: 0.5825
Epoch 00004: accuracy did not improve from 0.94873
64/78 [=======================>......] - ETA: 11:37 - loss: 0.3905 - accuracy: 0.9368 - mean_iou: 0.5814
Epoch 00004: accuracy did not improve from 0.94873
65/78 [========================>.....] - ETA: 10:47 - loss: 0.3904 - accuracy: 0.9369 - mean_iou: 0.5804
Epoch 00004: accuracy did not improve from 0.94873
66/78 [========================>.....] - ETA: 9:57 - loss: 0.3901 - accuracy: 0.9369 - mean_iou: 0.5789 
Epoch 00004: accuracy did not improve from 0.94873
67/78 [========================>.....] - ETA: 9:08 - loss: 0.3904 - accuracy: 0.9368 - mean_iou: 0.5779
Epoch 00004: accuracy did not improve from 0.94873
68/78 [=========================>....] - ETA: 8:17 - loss: 0.3912 - accuracy: 0.9366 - mean_iou: 0.5776
Epoch 00004: accuracy did not improve from 0.94873
69/78 [=========================>....] - ETA: 7:27 - loss: 0.3916 - accuracy: 0.9367 - mean_iou: 0.5780
Epoch 00004: accuracy did not improve from 0.94873
70/78 [=========================>....] - ETA: 6:37 - loss: 0.3912 - accuracy: 0.9369 - mean_iou: 0.5790
Epoch 00004: accuracy did not improve from 0.94873
71/78 [==========================>...] - ETA: 5:47 - loss: 0.3915 - accuracy: 0.9368 - mean_iou: 0.5802
Epoch 00004: accuracy did not improve from 0.94873
72/78 [==========================>...] - ETA: 4:57 - loss: 0.3921 - accuracy: 0.9365 - mean_iou: 0.5803
Epoch 00004: accuracy did not improve from 0.94873
73/78 [===========================>..] - ETA: 4:07 - loss: 0.3922 - accuracy: 0.9364 - mean_iou: 0.5796
Epoch 00004: accuracy did not improve from 0.94873
74/78 [===========================>..] - ETA: 3:18 - loss: 0.3914 - accuracy: 0.9365 - mean_iou: 0.5806
Epoch 00004: accuracy did not improve from 0.94873
75/78 [===========================>..] - ETA: 2:28 - loss: 0.3910 - accuracy: 0.9366 - mean_iou: 0.5803
Epoch 00004: accuracy did not improve from 0.94873
76/78 [============================>.] - ETA: 1:38 - loss: 0.3912 - accuracy: 0.9365 - mean_iou: 0.5790
Epoch 00004: accuracy did not improve from 0.94873
77/78 [============================>.] - ETA: 49s - loss: 0.3910 - accuracy: 0.9366 - mean_iou: 0.5806 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.

Epoch 00004: accuracy did not improve from 0.94873
78/78 [==============================] - ETA: 0s - loss: 0.3908 - accuracy: 0.9366 - mean_iou: 0.5795 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
78/78 [==============================] - 4139s 53s/step - loss: 0.3908 - accuracy: 0.9366 - mean_iou: 0.5795 - val_loss: 0.4269 - val_accuracy: 0.9307 - val_mean_iou: 0.4389 - lr: 9.6489e-04
Epoch 5/6
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.

Epoch 00005: accuracy did not improve from 0.94873
 1/78 [..............................] - ETA: 0s - loss: 0.4689 - accuracy: 0.9227 - mean_iou: 0.3598
Epoch 00005: accuracy did not improve from 0.94873
 2/78 [..............................] - ETA: 32:00 - loss: 0.4260 - accuracy: 0.9330 - mean_iou: 0.4379
Epoch 00005: accuracy did not improve from 0.94873
 3/78 [>.............................] - ETA: 42:01 - loss: 0.4156 - accuracy: 0.9312 - mean_iou: 0.4666
Epoch 00005: accuracy did not improve from 0.94873
 4/78 [>.............................] - ETA: 46:50 - loss: 0.4174 - accuracy: 0.9304 - mean_iou: 0.4891
Epoch 00005: accuracy did not improve from 0.94873
 5/78 [>.............................] - ETA: 49:21 - loss: 0.4069 - accuracy: 0.9319 - mean_iou: 0.5120
Epoch 00005: accuracy did not improve from 0.94873
 6/78 [=>............................] - ETA: 50:45 - loss: 0.4037 - accuracy: 0.9331 - mean_iou: 0.5114
Epoch 00005: accuracy did not improve from 0.94873
 7/78 [=>............................] - ETA: 51:27 - loss: 0.3981 - accuracy: 0.9360 - mean_iou: 0.5168
Epoch 00005: accuracy did not improve from 0.94873
 8/78 [==>...........................] - ETA: 51:49 - loss: 0.3999 - accuracy: 0.9361 - mean_iou: 0.5249
Epoch 00005: accuracy did not improve from 0.94873
 9/78 [==>...........................] - ETA: 51:55 - loss: 0.3969 - accuracy: 0.9379 - mean_iou: 0.5497
Epoch 00005: accuracy did not improve from 0.94873
10/78 [==>...........................] - ETA: 51:47 - loss: 0.3952 - accuracy: 0.9391 - mean_iou: 0.5599
Epoch 00005: accuracy did not improve from 0.94873
11/78 [===>..........................] - ETA: 51:34 - loss: 0.3866 - accuracy: 0.9411 - mean_iou: 0.5781
Epoch 00005: accuracy did not improve from 0.94873
12/78 [===>..........................] - ETA: 51:17 - loss: 0.3906 - accuracy: 0.9397 - mean_iou: 0.5752
Epoch 00005: accuracy did not improve from 0.94873
13/78 [====>.........................] - ETA: 50:53 - loss: 0.3903 - accuracy: 0.9391 - mean_iou: 0.5850
Epoch 00005: accuracy did not improve from 0.94873
14/78 [====>.........................] - ETA: 50:23 - loss: 0.3861 - accuracy: 0.9404 - mean_iou: 0.5933
Epoch 00005: accuracy did not improve from 0.94873
15/78 [====>.........................] - ETA: 49:51 - loss: 0.3865 - accuracy: 0.9400 - mean_iou: 0.5973
Epoch 00005: accuracy did not improve from 0.94873
16/78 [=====>........................] - ETA: 49:18 - loss: 0.3865 - accuracy: 0.9402 - mean_iou: 0.6021
Epoch 00005: accuracy did not improve from 0.94873
17/78 [=====>........................] - ETA: 48:42 - loss: 0.3849 - accuracy: 0.9403 - mean_iou: 0.6041
Epoch 00005: accuracy did not improve from 0.94873
18/78 [=====>........................] - ETA: 48:04 - loss: 0.3851 - accuracy: 0.9395 - mean_iou: 0.6016
Epoch 00005: accuracy did not improve from 0.94873
19/78 [======>.......................] - ETA: 47:25 - loss: 0.3846 - accuracy: 0.9392 - mean_iou: 0.6014
Epoch 00005: accuracy did not improve from 0.94873
20/78 [======>.......................] - ETA: 46:44 - loss: 0.3847 - accuracy: 0.9388 - mean_iou: 0.6020
Epoch 00005: accuracy did not improve from 0.94873
21/78 [=======>......................] - ETA: 46:02 - loss: 0.3832 - accuracy: 0.9388 - mean_iou: 0.6011
Epoch 00005: accuracy did not improve from 0.94873
22/78 [=======>......................] - ETA: 45:19 - loss: 0.3811 - accuracy: 0.9389 - mean_iou: 0.5963
Epoch 00005: accuracy did not improve from 0.94873
23/78 [=======>......................] - ETA: 44:37 - loss: 0.3813 - accuracy: 0.9388 - mean_iou: 0.5936
Epoch 00005: accuracy did not improve from 0.94873
24/78 [========>.....................] - ETA: 43:52 - loss: 0.3804 - accuracy: 0.9386 - mean_iou: 0.5971
Epoch 00005: accuracy did not improve from 0.94873
25/78 [========>.....................] - ETA: 43:09 - loss: 0.3801 - accuracy: 0.9384 - mean_iou: 0.5952
Epoch 00005: accuracy did not improve from 0.94873
26/78 [=========>....................] - ETA: 42:23 - loss: 0.3809 - accuracy: 0.9382 - mean_iou: 0.5929
Epoch 00005: accuracy did not improve from 0.94873
27/78 [=========>....................] - ETA: 41:38 - loss: 0.3803 - accuracy: 0.9383 - mean_iou: 0.5921
Epoch 00005: accuracy did not improve from 0.94873
28/78 [=========>....................] - ETA: 40:54 - loss: 0.3805 - accuracy: 0.9382 - mean_iou: 0.5900
Epoch 00005: accuracy did not improve from 0.94873
29/78 [==========>...................] - ETA: 40:07 - loss: 0.3809 - accuracy: 0.9385 - mean_iou: 0.5912
Epoch 00005: accuracy did not improve from 0.94873
30/78 [==========>...................] - ETA: 39:21 - loss: 0.3822 - accuracy: 0.9381 - mean_iou: 0.5885
Epoch 00005: accuracy did not improve from 0.94873
31/78 [==========>...................] - ETA: 38:35 - loss: 0.3825 - accuracy: 0.9380 - mean_iou: 0.5882
Epoch 00005: accuracy did not improve from 0.94873
32/78 [===========>..................] - ETA: 37:48 - loss: 0.3821 - accuracy: 0.9381 - mean_iou: 0.5868
Epoch 00005: accuracy did not improve from 0.94873
33/78 [===========>..................] - ETA: 37:01 - loss: 0.3822 - accuracy: 0.9379 - mean_iou: 0.5868
Epoch 00005: accuracy did not improve from 0.94873
34/78 [============>.................] - ETA: 36:13 - loss: 0.3817 - accuracy: 0.9379 - mean_iou: 0.5885
Epoch 00005: accuracy did not improve from 0.94873
35/78 [============>.................] - ETA: 35:25 - loss: 0.3809 - accuracy: 0.9380 - mean_iou: 0.5889
Epoch 00005: accuracy did not improve from 0.94873
36/78 [============>.................] - ETA: 34:37 - loss: 0.3812 - accuracy: 0.9383 - mean_iou: 0.5884
Epoch 00005: accuracy did not improve from 0.94873
37/78 [=============>................] - ETA: 33:49 - loss: 0.3815 - accuracy: 0.9382 - mean_iou: 0.5889
Epoch 00005: accuracy did not improve from 0.94873
38/78 [=============>................] - ETA: 33:01 - loss: 0.3833 - accuracy: 0.9380 - mean_iou: 0.5839
Epoch 00005: accuracy did not improve from 0.94873
39/78 [==============>...............] - ETA: 32:12 - loss: 0.3825 - accuracy: 0.9383 - mean_iou: 0.5849
Epoch 00005: accuracy did not improve from 0.94873
40/78 [==============>...............] - ETA: 31:24 - loss: 0.3835 - accuracy: 0.9385 - mean_iou: 0.5842
Epoch 00005: accuracy did not improve from 0.94873
41/78 [==============>...............] - ETA: 30:35 - loss: 0.3837 - accuracy: 0.9386 - mean_iou: 0.5851
Epoch 00005: accuracy did not improve from 0.94873
42/78 [===============>..............] - ETA: 29:46 - loss: 0.3830 - accuracy: 0.9387 - mean_iou: 0.5859
Epoch 00005: accuracy did not improve from 0.94873
43/78 [===============>..............] - ETA: 28:57 - loss: 0.3823 - accuracy: 0.9391 - mean_iou: 0.5869
Epoch 00005: accuracy did not improve from 0.94873
44/78 [===============>..............] - ETA: 28:08 - loss: 0.3859 - accuracy: 0.9377 - mean_iou: 0.5850
Epoch 00005: accuracy did not improve from 0.94873
45/78 [================>.............] - ETA: 27:19 - loss: 0.3869 - accuracy: 0.9376 - mean_iou: 0.5848
Epoch 00005: accuracy did not improve from 0.94873
46/78 [================>.............] - ETA: 26:31 - loss: 0.3864 - accuracy: 0.9376 - mean_iou: 0.5857
Epoch 00005: accuracy did not improve from 0.94873
47/78 [=================>............] - ETA: 25:41 - loss: 0.3859 - accuracy: 0.9377 - mean_iou: 0.5869
Epoch 00005: accuracy did not improve from 0.94873
48/78 [=================>............] - ETA: 24:52 - loss: 0.3856 - accuracy: 0.9377 - mean_iou: 0.5852
Epoch 00005: accuracy did not improve from 0.94873
49/78 [=================>............] - ETA: 24:03 - loss: 0.3852 - accuracy: 0.9377 - mean_iou: 0.5847
Epoch 00005: accuracy did not improve from 0.94873
50/78 [==================>...........] - ETA: 23:14 - loss: 0.3861 - accuracy: 0.9373 - mean_iou: 0.5817
Epoch 00005: accuracy did not improve from 0.94873
51/78 [==================>...........] - ETA: 22:24 - loss: 0.3865 - accuracy: 0.9373 - mean_iou: 0.5808
Epoch 00005: accuracy did not improve from 0.94873
52/78 [===================>..........] - ETA: 21:35 - loss: 0.3863 - accuracy: 0.9376 - mean_iou: 0.5805
Epoch 00005: accuracy did not improve from 0.94873
53/78 [===================>..........] - ETA: 20:46 - loss: 0.3867 - accuracy: 0.9374 - mean_iou: 0.5797
Epoch 00005: accuracy did not improve from 0.94873
54/78 [===================>..........] - ETA: 19:56 - loss: 0.3862 - accuracy: 0.9376 - mean_iou: 0.5801
Epoch 00005: accuracy did not improve from 0.94873
55/78 [====================>.........] - ETA: 19:07 - loss: 0.3857 - accuracy: 0.9377 - mean_iou: 0.5797
Epoch 00005: accuracy did not improve from 0.94873
56/78 [====================>.........] - ETA: 18:17 - loss: 0.3864 - accuracy: 0.9376 - mean_iou: 0.5793
Epoch 00005: accuracy did not improve from 0.94873
57/78 [====================>.........] - ETA: 17:28 - loss: 0.3862 - accuracy: 0.9378 - mean_iou: 0.5789
Epoch 00005: accuracy did not improve from 0.94873
58/78 [=====================>........] - ETA: 16:38 - loss: 0.3864 - accuracy: 0.9376 - mean_iou: 0.5785
Epoch 00005: accuracy did not improve from 0.94873
59/78 [=====================>........] - ETA: 15:49 - loss: 0.3869 - accuracy: 0.9377 - mean_iou: 0.5779
Epoch 00005: accuracy did not improve from 0.94873
60/78 [======================>.......] - ETA: 14:59 - loss: 0.3878 - accuracy: 0.9375 - mean_iou: 0.5774
Epoch 00005: accuracy did not improve from 0.94873
61/78 [======================>.......] - ETA: 14:09 - loss: 0.3871 - accuracy: 0.9376 - mean_iou: 0.5781
Epoch 00005: accuracy did not improve from 0.94873
62/78 [======================>.......] - ETA: 13:19 - loss: 0.3865 - accuracy: 0.9378 - mean_iou: 0.5803
Epoch 00005: accuracy did not improve from 0.94873
63/78 [=======================>......] - ETA: 12:30 - loss: 0.3865 - accuracy: 0.9378 - mean_iou: 0.5798
Epoch 00005: accuracy did not improve from 0.94873
64/78 [=======================>......] - ETA: 11:40 - loss: 0.3868 - accuracy: 0.9377 - mean_iou: 0.5782
Epoch 00005: accuracy did not improve from 0.94873
65/78 [========================>.....] - ETA: 10:50 - loss: 0.3859 - accuracy: 0.9379 - mean_iou: 0.5799
Epoch 00005: accuracy did not improve from 0.94873
66/78 [========================>.....] - ETA: 10:00 - loss: 0.3852 - accuracy: 0.9380 - mean_iou: 0.5818
Epoch 00005: accuracy did not improve from 0.94873
67/78 [========================>.....] - ETA: 9:10 - loss: 0.3852 - accuracy: 0.9382 - mean_iou: 0.5817 
Epoch 00005: accuracy did not improve from 0.94873
68/78 [=========================>....] - ETA: 8:20 - loss: 0.3853 - accuracy: 0.9382 - mean_iou: 0.5816
Epoch 00005: accuracy did not improve from 0.94873
69/78 [=========================>....] - ETA: 7:29 - loss: 0.3847 - accuracy: 0.9383 - mean_iou: 0.5836
Epoch 00005: accuracy did not improve from 0.94873
70/78 [=========================>....] - ETA: 6:39 - loss: 0.3851 - accuracy: 0.9382 - mean_iou: 0.5830
Epoch 00005: accuracy did not improve from 0.94873
71/78 [==========================>...] - ETA: 5:49 - loss: 0.3854 - accuracy: 0.9382 - mean_iou: 0.5829
Epoch 00005: accuracy did not improve from 0.94873
72/78 [==========================>...] - ETA: 4:59 - loss: 0.3855 - accuracy: 0.9383 - mean_iou: 0.5828
Epoch 00005: accuracy did not improve from 0.94873
73/78 [===========================>..] - ETA: 4:09 - loss: 0.3859 - accuracy: 0.9381 - mean_iou: 0.5824
Epoch 00005: accuracy did not improve from 0.94873
74/78 [===========================>..] - ETA: 3:19 - loss: 0.3861 - accuracy: 0.9380 - mean_iou: 0.5830
Epoch 00005: accuracy did not improve from 0.94873
75/78 [===========================>..] - ETA: 2:29 - loss: 0.3849 - accuracy: 0.9382 - mean_iou: 0.5834
Epoch 00005: accuracy did not improve from 0.94873
76/78 [============================>.] - ETA: 1:39 - loss: 0.3846 - accuracy: 0.9382 - mean_iou: 0.5833
Epoch 00005: accuracy did not improve from 0.94873
77/78 [============================>.] - ETA: 49s - loss: 0.3846 - accuracy: 0.9383 - mean_iou: 0.5828 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.

Epoch 00005: accuracy did not improve from 0.94873
78/78 [==============================] - ETA: 0s - loss: 0.3852 - accuracy: 0.9380 - mean_iou: 0.5819 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
78/78 [==============================] - 4161s 53s/step - loss: 0.3852 - accuracy: 0.9380 - mean_iou: 0.5819 - val_loss: 0.4410 - val_accuracy: 0.9433 - val_mean_iou: 0.5734 - lr: 9.3815e-04
Epoch 6/6
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.

Epoch 00006: accuracy did not improve from 0.94873
 1/78 [..............................] - ETA: 0s - loss: 0.3951 - accuracy: 0.9246 - mean_iou: 0.5942
Epoch 00006: accuracy did not improve from 0.94873
 2/78 [..............................] - ETA: 32:10 - loss: 0.3942 - accuracy: 0.9266 - mean_iou: 0.5603
Epoch 00006: accuracy did not improve from 0.94873
 3/78 [>.............................] - ETA: 42:20 - loss: 0.3972 - accuracy: 0.9276 - mean_iou: 0.5554
Epoch 00006: accuracy did not improve from 0.94873
 4/78 [>.............................] - ETA: 46:56 - loss: 0.3925 - accuracy: 0.9290 - mean_iou: 0.5520
Epoch 00006: accuracy did not improve from 0.94873
 5/78 [>.............................] - ETA: 49:28 - loss: 0.3934 - accuracy: 0.9283 - mean_iou: 0.5256
Epoch 00006: accuracy did not improve from 0.94873
 6/78 [=>............................] - ETA: 50:48 - loss: 0.3918 - accuracy: 0.9286 - mean_iou: 0.5168
Epoch 00006: accuracy did not improve from 0.94873
 7/78 [=>............................] - ETA: 51:31 - loss: 0.3937 - accuracy: 0.9284 - mean_iou: 0.5118
Epoch 00006: accuracy did not improve from 0.94873
 8/78 [==>...........................] - ETA: 51:51 - loss: 0.3898 - accuracy: 0.9303 - mean_iou: 0.5199
Epoch 00006: accuracy did not improve from 0.94873
 9/78 [==>...........................] - ETA: 51:52 - loss: 0.3839 - accuracy: 0.9320 - mean_iou: 0.5310
Epoch 00006: accuracy did not improve from 0.94873
10/78 [==>...........................] - ETA: 51:49 - loss: 0.3830 - accuracy: 0.9316 - mean_iou: 0.5382
Epoch 00006: accuracy did not improve from 0.94873
11/78 [===>..........................] - ETA: 51:33 - loss: 0.3860 - accuracy: 0.9331 - mean_iou: 0.5514
Epoch 00006: accuracy did not improve from 0.94873
12/78 [===>..........................] - ETA: 51:11 - loss: 0.3838 - accuracy: 0.9342 - mean_iou: 0.5548
Epoch 00006: accuracy did not improve from 0.94873
13/78 [====>.........................] - ETA: 50:48 - loss: 0.3824 - accuracy: 0.9347 - mean_iou: 0.5575
Epoch 00006: accuracy did not improve from 0.94873
14/78 [====>.........................] - ETA: 50:20 - loss: 0.3800 - accuracy: 0.9358 - mean_iou: 0.5644
Epoch 00006: accuracy did not improve from 0.94873
15/78 [====>.........................] - ETA: 49:51 - loss: 0.3826 - accuracy: 0.9353 - mean_iou: 0.5645
Epoch 00006: accuracy did not improve from 0.94873
16/78 [=====>........................] - ETA: 49:17 - loss: 0.3796 - accuracy: 0.9358 - mean_iou: 0.5726
Epoch 00006: accuracy did not improve from 0.94873
17/78 [=====>........................] - ETA: 48:42 - loss: 0.3841 - accuracy: 0.9348 - mean_iou: 0.5711
Epoch 00006: accuracy did not improve from 0.94873
18/78 [=====>........................] - ETA: 48:06 - loss: 0.3861 - accuracy: 0.9354 - mean_iou: 0.5700
Epoch 00006: accuracy did not improve from 0.94873
19/78 [======>.......................] - ETA: 47:29 - loss: 0.3826 - accuracy: 0.9362 - mean_iou: 0.5735
Epoch 00006: accuracy did not improve from 0.94873
20/78 [======>.......................] - ETA: 46:49 - loss: 0.3830 - accuracy: 0.9356 - mean_iou: 0.5727
Epoch 00006: accuracy did not improve from 0.94873
21/78 [=======>......................] - ETA: 46:07 - loss: 0.3834 - accuracy: 0.9355 - mean_iou: 0.5699
Epoch 00006: accuracy did not improve from 0.94873
22/78 [=======>......................] - ETA: 45:26 - loss: 0.3826 - accuracy: 0.9355 - mean_iou: 0.5683
Epoch 00006: accuracy did not improve from 0.94873
23/78 [=======>......................] - ETA: 44:42 - loss: 0.3854 - accuracy: 0.9355 - mean_iou: 0.5620
Epoch 00006: accuracy did not improve from 0.94873
24/78 [========>.....................] - ETA: 44:00 - loss: 0.3841 - accuracy: 0.9361 - mean_iou: 0.5656
Epoch 00006: accuracy did not improve from 0.94873
25/78 [========>.....................] - ETA: 43:15 - loss: 0.3844 - accuracy: 0.9360 - mean_iou: 0.5656
Epoch 00006: accuracy did not improve from 0.94873
26/78 [=========>....................] - ETA: 42:30 - loss: 0.3856 - accuracy: 0.9361 - mean_iou: 0.5635
Epoch 00006: accuracy did not improve from 0.94873
27/78 [=========>....................] - ETA: 41:44 - loss: 0.3837 - accuracy: 0.9364 - mean_iou: 0.5679
Epoch 00006: accuracy did not improve from 0.94873
28/78 [=========>....................] - ETA: 40:58 - loss: 0.3844 - accuracy: 0.9367 - mean_iou: 0.5666
Epoch 00006: accuracy did not improve from 0.94873
29/78 [==========>...................] - ETA: 40:12 - loss: 0.3841 - accuracy: 0.9369 - mean_iou: 0.5678
Epoch 00006: accuracy did not improve from 0.94873
30/78 [==========>...................] - ETA: 39:25 - loss: 0.3844 - accuracy: 0.9365 - mean_iou: 0.5678
Epoch 00006: accuracy did not improve from 0.94873
31/78 [==========>...................] - ETA: 38:38 - loss: 0.3838 - accuracy: 0.9367 - mean_iou: 0.5696
Epoch 00006: accuracy did not improve from 0.94873
32/78 [===========>..................] - ETA: 37:51 - loss: 0.3830 - accuracy: 0.9370 - mean_iou: 0.5712
Epoch 00006: accuracy did not improve from 0.94873
33/78 [===========>..................] - ETA: 37:03 - loss: 0.3830 - accuracy: 0.9370 - mean_iou: 0.5719
Epoch 00006: accuracy did not improve from 0.94873
34/78 [============>.................] - ETA: 36:16 - loss: 0.3838 - accuracy: 0.9367 - mean_iou: 0.5727
Epoch 00006: accuracy did not improve from 0.94873
35/78 [============>.................] - ETA: 35:28 - loss: 0.3816 - accuracy: 0.9373 - mean_iou: 0.5757
Epoch 00006: accuracy did not improve from 0.94873
36/78 [============>.................] - ETA: 34:40 - loss: 0.3812 - accuracy: 0.9373 - mean_iou: 0.5782
Epoch 00006: accuracy did not improve from 0.94873
37/78 [=============>................] - ETA: 33:52 - loss: 0.3805 - accuracy: 0.9373 - mean_iou: 0.5800
Epoch 00006: accuracy did not improve from 0.94873
38/78 [=============>................] - ETA: 33:04 - loss: 0.3795 - accuracy: 0.9376 - mean_iou: 0.5824
Epoch 00006: accuracy did not improve from 0.94873
39/78 [==============>...............] - ETA: 32:15 - loss: 0.3811 - accuracy: 0.9374 - mean_iou: 0.5821
Epoch 00006: accuracy did not improve from 0.94873
40/78 [==============>...............] - ETA: 31:27 - loss: 0.3802 - accuracy: 0.9375 - mean_iou: 0.5823
Epoch 00006: accuracy did not improve from 0.94873
41/78 [==============>...............] - ETA: 30:38 - loss: 0.3793 - accuracy: 0.9378 - mean_iou: 0.5839
Epoch 00006: accuracy did not improve from 0.94873
42/78 [===============>..............] - ETA: 29:50 - loss: 0.3788 - accuracy: 0.9379 - mean_iou: 0.5855
Epoch 00006: accuracy did not improve from 0.94873
43/78 [===============>..............] - ETA: 29:01 - loss: 0.3781 - accuracy: 0.9383 - mean_iou: 0.5866
Epoch 00006: accuracy did not improve from 0.94873
44/78 [===============>..............] - ETA: 28:12 - loss: 0.3778 - accuracy: 0.9385 - mean_iou: 0.5878
Epoch 00006: accuracy did not improve from 0.94873
45/78 [================>.............] - ETA: 27:23 - loss: 0.3769 - accuracy: 0.9388 - mean_iou: 0.5874
Epoch 00006: accuracy did not improve from 0.94873
46/78 [================>.............] - ETA: 26:34 - loss: 0.3762 - accuracy: 0.9390 - mean_iou: 0.5886
Epoch 00006: accuracy did not improve from 0.94873
47/78 [=================>............] - ETA: 25:45 - loss: 0.3748 - accuracy: 0.9395 - mean_iou: 0.5918
Epoch 00006: accuracy did not improve from 0.94873
48/78 [=================>............] - ETA: 24:55 - loss: 0.3742 - accuracy: 0.9396 - mean_iou: 0.5927
Epoch 00006: accuracy did not improve from 0.94873
49/78 [=================>............] - ETA: 24:06 - loss: 0.3753 - accuracy: 0.9394 - mean_iou: 0.5915
Epoch 00006: accuracy did not improve from 0.94873
50/78 [==================>...........] - ETA: 23:17 - loss: 0.3745 - accuracy: 0.9396 - mean_iou: 0.5935
Epoch 00006: accuracy did not improve from 0.94873
51/78 [==================>...........] - ETA: 22:27 - loss: 0.3740 - accuracy: 0.9398 - mean_iou: 0.5947
Epoch 00006: accuracy did not improve from 0.94873
52/78 [===================>..........] - ETA: 21:38 - loss: 0.3744 - accuracy: 0.9401 - mean_iou: 0.5967
Epoch 00006: accuracy did not improve from 0.94873
53/78 [===================>..........] - ETA: 20:48 - loss: 0.3743 - accuracy: 0.9401 - mean_iou: 0.5954
Epoch 00006: accuracy did not improve from 0.94873
54/78 [===================>..........] - ETA: 19:59 - loss: 0.3744 - accuracy: 0.9401 - mean_iou: 0.5971
Epoch 00006: accuracy did not improve from 0.94873
55/78 [====================>.........] - ETA: 19:09 - loss: 0.3757 - accuracy: 0.9398 - mean_iou: 0.5942
Epoch 00006: accuracy did not improve from 0.94873
56/78 [====================>.........] - ETA: 18:19 - loss: 0.3751 - accuracy: 0.9400 - mean_iou: 0.5942
Epoch 00006: accuracy did not improve from 0.94873
57/78 [====================>.........] - ETA: 17:30 - loss: 0.3758 - accuracy: 0.9399 - mean_iou: 0.5932
Epoch 00006: accuracy did not improve from 0.94873
58/78 [=====================>........] - ETA: 16:40 - loss: 0.3755 - accuracy: 0.9401 - mean_iou: 0.5935
Epoch 00006: accuracy did not improve from 0.94873
59/78 [=====================>........] - ETA: 15:50 - loss: 0.3756 - accuracy: 0.9401 - mean_iou: 0.5934
Epoch 00006: accuracy did not improve from 0.94873
60/78 [======================>.......] - ETA: 15:00 - loss: 0.3757 - accuracy: 0.9401 - mean_iou: 0.5934
Epoch 00006: accuracy did not improve from 0.94873
61/78 [======================>.......] - ETA: 14:10 - loss: 0.3757 - accuracy: 0.9400 - mean_iou: 0.5941
Epoch 00006: accuracy did not improve from 0.94873
62/78 [======================>.......] - ETA: 13:20 - loss: 0.3753 - accuracy: 0.9402 - mean_iou: 0.5947
Epoch 00006: accuracy did not improve from 0.94873
63/78 [=======================>......] - ETA: 12:30 - loss: 0.3759 - accuracy: 0.9402 - mean_iou: 0.5942
Epoch 00006: accuracy did not improve from 0.94873
64/78 [=======================>......] - ETA: 11:40 - loss: 0.3762 - accuracy: 0.9401 - mean_iou: 0.5931
Epoch 00006: accuracy did not improve from 0.94873
65/78 [========================>.....] - ETA: 10:50 - loss: 0.3754 - accuracy: 0.9402 - mean_iou: 0.5937
Epoch 00006: accuracy did not improve from 0.94873
66/78 [========================>.....] - ETA: 10:00 - loss: 0.3762 - accuracy: 0.9401 - mean_iou: 0.5936
Epoch 00006: accuracy did not improve from 0.94873
67/78 [========================>.....] - ETA: 9:11 - loss: 0.3761 - accuracy: 0.9399 - mean_iou: 0.5941 
Epoch 00006: accuracy did not improve from 0.94873
68/78 [=========================>....] - ETA: 8:20 - loss: 0.3762 - accuracy: 0.9400 - mean_iou: 0.5943
Epoch 00006: accuracy did not improve from 0.94873
69/78 [=========================>....] - ETA: 7:29 - loss: 0.3761 - accuracy: 0.9401 - mean_iou: 0.5939
Epoch 00006: accuracy did not improve from 0.94873
70/78 [=========================>....] - ETA: 6:39 - loss: 0.3753 - accuracy: 0.9401 - mean_iou: 0.5939
Epoch 00006: accuracy did not improve from 0.94873
71/78 [==========================>...] - ETA: 5:49 - loss: 0.3755 - accuracy: 0.9400 - mean_iou: 0.5923
Epoch 00006: accuracy did not improve from 0.94873
72/78 [==========================>...] - ETA: 4:59 - loss: 0.3762 - accuracy: 0.9401 - mean_iou: 0.5916
Epoch 00006: accuracy did not improve from 0.94873
73/78 [===========================>..] - ETA: 4:09 - loss: 0.3764 - accuracy: 0.9399 - mean_iou: 0.5914
Epoch 00006: accuracy did not improve from 0.94873
74/78 [===========================>..] - ETA: 3:19 - loss: 0.3759 - accuracy: 0.9400 - mean_iou: 0.5920
Epoch 00006: accuracy did not improve from 0.94873
75/78 [===========================>..] - ETA: 2:29 - loss: 0.3756 - accuracy: 0.9400 - mean_iou: 0.5927
Epoch 00006: accuracy did not improve from 0.94873
76/78 [============================>.] - ETA: 1:39 - loss: 0.3753 - accuracy: 0.9400 - mean_iou: 0.5940
Epoch 00006: accuracy did not improve from 0.94873
77/78 [============================>.] - ETA: 49s - loss: 0.3748 - accuracy: 0.9402 - mean_iou: 0.5948 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.

Epoch 00006: accuracy did not improve from 0.94873
78/78 [==============================] - ETA: 0s - loss: 0.3737 - accuracy: 0.9405 - mean_iou: 0.5968 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
78/78 [==============================] - 4173s 54s/step - loss: 0.3737 - accuracy: 0.9405 - mean_iou: 0.5968 - val_loss: 0.4118 - val_accuracy: 0.9390 - val_mean_iou: 0.5383 - lr: 9.0451e-04

model_json = model.to_json()
with open("model4000_6.json", "w") as json_file:
    json_file.write(model_json)

model.save_weights("model4000_6.h5")

plt.figure(figsize=(12,4))
plt.subplot(131)
plt.plot(history.epoch, history.history["loss"], label="Train loss")
plt.plot(history.epoch, history.history["val_loss"], label="Valid loss")
plt.legend()
plt.subplot(132)
plt.plot(history.epoch, history.history["accuracy"], label="Train accuracy")
plt.plot(history.epoch, history.history["val_accuracy"], label="Valid accuracy")
plt.legend()
plt.subplot(133)
plt.plot(history.epoch, history.history["mean_iou"], label="Train iou")
plt.plot(history.epoch, history.history["val_mean_iou"], label="Valid iou")
plt.legend()
plt.show()

model.summary()

Model: "model"
__________________________________________________________________________________________________
Layer (type)                    Output Shape         Param #     Connected to                     
==================================================================================================
input_1 (InputLayer)            [(None, 256, 256, 1) 0                                            
__________________________________________________________________________________________________
conv2d (Conv2D)                 (None, 256, 256, 32) 288         input_1[0][0]                    
__________________________________________________________________________________________________
batch_normalization (BatchNorma (None, 256, 256, 32) 128         conv2d[0][0]                     
__________________________________________________________________________________________________
leaky_re_lu (LeakyReLU)         (None, 256, 256, 32) 0           batch_normalization[0][0]        
__________________________________________________________________________________________________
conv2d_1 (Conv2D)               (None, 256, 256, 64) 2048        leaky_re_lu[0][0]                
__________________________________________________________________________________________________
max_pooling2d (MaxPooling2D)    (None, 128, 128, 64) 0           conv2d_1[0][0]                   
__________________________________________________________________________________________________
batch_normalization_1 (BatchNor (None, 128, 128, 64) 256         max_pooling2d[0][0]              
__________________________________________________________________________________________________
leaky_re_lu_1 (LeakyReLU)       (None, 128, 128, 64) 0           batch_normalization_1[0][0]      
__________________________________________________________________________________________________
conv2d_2 (Conv2D)               (None, 128, 128, 64) 36864       leaky_re_lu_1[0][0]              
__________________________________________________________________________________________________
batch_normalization_2 (BatchNor (None, 128, 128, 64) 256         conv2d_2[0][0]                   
__________________________________________________________________________________________________
leaky_re_lu_2 (LeakyReLU)       (None, 128, 128, 64) 0           batch_normalization_2[0][0]      
__________________________________________________________________________________________________
conv2d_3 (Conv2D)               (None, 128, 128, 64) 36864       leaky_re_lu_2[0][0]              
__________________________________________________________________________________________________
add (Add)                       (None, 128, 128, 64) 0           conv2d_3[0][0]                   
                                                                 max_pooling2d[0][0]              
__________________________________________________________________________________________________
batch_normalization_3 (BatchNor (None, 128, 128, 64) 256         add[0][0]                        
__________________________________________________________________________________________________
leaky_re_lu_3 (LeakyReLU)       (None, 128, 128, 64) 0           batch_normalization_3[0][0]      
__________________________________________________________________________________________________
conv2d_4 (Conv2D)               (None, 128, 128, 64) 36864       leaky_re_lu_3[0][0]              
__________________________________________________________________________________________________
batch_normalization_4 (BatchNor (None, 128, 128, 64) 256         conv2d_4[0][0]                   
__________________________________________________________________________________________________
leaky_re_lu_4 (LeakyReLU)       (None, 128, 128, 64) 0           batch_normalization_4[0][0]      
__________________________________________________________________________________________________
conv2d_5 (Conv2D)               (None, 128, 128, 64) 36864       leaky_re_lu_4[0][0]              
__________________________________________________________________________________________________
add_1 (Add)                     (None, 128, 128, 64) 0           conv2d_5[0][0]                   
                                                                 add[0][0]                        
__________________________________________________________________________________________________
batch_normalization_5 (BatchNor (None, 128, 128, 64) 256         add_1[0][0]                      
__________________________________________________________________________________________________
leaky_re_lu_5 (LeakyReLU)       (None, 128, 128, 64) 0           batch_normalization_5[0][0]      
__________________________________________________________________________________________________
conv2d_6 (Conv2D)               (None, 128, 128, 128 8192        leaky_re_lu_5[0][0]              
__________________________________________________________________________________________________
max_pooling2d_1 (MaxPooling2D)  (None, 64, 64, 128)  0           conv2d_6[0][0]                   
__________________________________________________________________________________________________
batch_normalization_6 (BatchNor (None, 64, 64, 128)  512         max_pooling2d_1[0][0]            
__________________________________________________________________________________________________
leaky_re_lu_6 (LeakyReLU)       (None, 64, 64, 128)  0           batch_normalization_6[0][0]      
__________________________________________________________________________________________________
conv2d_7 (Conv2D)               (None, 64, 64, 128)  147456      leaky_re_lu_6[0][0]              
__________________________________________________________________________________________________
batch_normalization_7 (BatchNor (None, 64, 64, 128)  512         conv2d_7[0][0]                   
__________________________________________________________________________________________________
leaky_re_lu_7 (LeakyReLU)       (None, 64, 64, 128)  0           batch_normalization_7[0][0]      
__________________________________________________________________________________________________
conv2d_8 (Conv2D)               (None, 64, 64, 128)  147456      leaky_re_lu_7[0][0]              
__________________________________________________________________________________________________
add_2 (Add)                     (None, 64, 64, 128)  0           conv2d_8[0][0]                   
                                                                 max_pooling2d_1[0][0]            
__________________________________________________________________________________________________
batch_normalization_8 (BatchNor (None, 64, 64, 128)  512         add_2[0][0]                      
__________________________________________________________________________________________________
leaky_re_lu_8 (LeakyReLU)       (None, 64, 64, 128)  0           batch_normalization_8[0][0]      
__________________________________________________________________________________________________
conv2d_9 (Conv2D)               (None, 64, 64, 128)  147456      leaky_re_lu_8[0][0]              
__________________________________________________________________________________________________
batch_normalization_9 (BatchNor (None, 64, 64, 128)  512         conv2d_9[0][0]                   
__________________________________________________________________________________________________
leaky_re_lu_9 (LeakyReLU)       (None, 64, 64, 128)  0           batch_normalization_9[0][0]      
__________________________________________________________________________________________________
conv2d_10 (Conv2D)              (None, 64, 64, 128)  147456      leaky_re_lu_9[0][0]              
__________________________________________________________________________________________________
add_3 (Add)                     (None, 64, 64, 128)  0           conv2d_10[0][0]                  
                                                                 add_2[0][0]                      
__________________________________________________________________________________________________
batch_normalization_10 (BatchNo (None, 64, 64, 128)  512         add_3[0][0]                      
__________________________________________________________________________________________________
leaky_re_lu_10 (LeakyReLU)      (None, 64, 64, 128)  0           batch_normalization_10[0][0]     
__________________________________________________________________________________________________
conv2d_11 (Conv2D)              (None, 64, 64, 256)  32768       leaky_re_lu_10[0][0]             
__________________________________________________________________________________________________
max_pooling2d_2 (MaxPooling2D)  (None, 32, 32, 256)  0           conv2d_11[0][0]                  
__________________________________________________________________________________________________
batch_normalization_11 (BatchNo (None, 32, 32, 256)  1024        max_pooling2d_2[0][0]            
__________________________________________________________________________________________________
leaky_re_lu_11 (LeakyReLU)      (None, 32, 32, 256)  0           batch_normalization_11[0][0]     
__________________________________________________________________________________________________
conv2d_12 (Conv2D)              (None, 32, 32, 256)  589824      leaky_re_lu_11[0][0]             
__________________________________________________________________________________________________
batch_normalization_12 (BatchNo (None, 32, 32, 256)  1024        conv2d_12[0][0]                  
__________________________________________________________________________________________________
leaky_re_lu_12 (LeakyReLU)      (None, 32, 32, 256)  0           batch_normalization_12[0][0]     
__________________________________________________________________________________________________
conv2d_13 (Conv2D)              (None, 32, 32, 256)  589824      leaky_re_lu_12[0][0]             
__________________________________________________________________________________________________
add_4 (Add)                     (None, 32, 32, 256)  0           conv2d_13[0][0]                  
                                                                 max_pooling2d_2[0][0]            
__________________________________________________________________________________________________
batch_normalization_13 (BatchNo (None, 32, 32, 256)  1024        add_4[0][0]                      
__________________________________________________________________________________________________
leaky_re_lu_13 (LeakyReLU)      (None, 32, 32, 256)  0           batch_normalization_13[0][0]     
__________________________________________________________________________________________________
conv2d_14 (Conv2D)              (None, 32, 32, 256)  589824      leaky_re_lu_13[0][0]             
__________________________________________________________________________________________________
batch_normalization_14 (BatchNo (None, 32, 32, 256)  1024        conv2d_14[0][0]                  
__________________________________________________________________________________________________
leaky_re_lu_14 (LeakyReLU)      (None, 32, 32, 256)  0           batch_normalization_14[0][0]     
__________________________________________________________________________________________________
conv2d_15 (Conv2D)              (None, 32, 32, 256)  589824      leaky_re_lu_14[0][0]             
__________________________________________________________________________________________________
add_5 (Add)                     (None, 32, 32, 256)  0           conv2d_15[0][0]                  
                                                                 add_4[0][0]                      
__________________________________________________________________________________________________
batch_normalization_15 (BatchNo (None, 32, 32, 256)  1024        add_5[0][0]                      
__________________________________________________________________________________________________
leaky_re_lu_15 (LeakyReLU)      (None, 32, 32, 256)  0           batch_normalization_15[0][0]     
__________________________________________________________________________________________________
conv2d_16 (Conv2D)              (None, 32, 32, 512)  131072      leaky_re_lu_15[0][0]             
__________________________________________________________________________________________________
max_pooling2d_3 (MaxPooling2D)  (None, 16, 16, 512)  0           conv2d_16[0][0]                  
__________________________________________________________________________________________________
batch_normalization_16 (BatchNo (None, 16, 16, 512)  2048        max_pooling2d_3[0][0]            
__________________________________________________________________________________________________
leaky_re_lu_16 (LeakyReLU)      (None, 16, 16, 512)  0           batch_normalization_16[0][0]     
__________________________________________________________________________________________________
conv2d_17 (Conv2D)              (None, 16, 16, 512)  2359296     leaky_re_lu_16[0][0]             
__________________________________________________________________________________________________
batch_normalization_17 (BatchNo (None, 16, 16, 512)  2048        conv2d_17[0][0]                  
__________________________________________________________________________________________________
leaky_re_lu_17 (LeakyReLU)      (None, 16, 16, 512)  0           batch_normalization_17[0][0]     
__________________________________________________________________________________________________
conv2d_18 (Conv2D)              (None, 16, 16, 512)  2359296     leaky_re_lu_17[0][0]             
__________________________________________________________________________________________________
add_6 (Add)                     (None, 16, 16, 512)  0           conv2d_18[0][0]                  
                                                                 max_pooling2d_3[0][0]            
__________________________________________________________________________________________________
batch_normalization_18 (BatchNo (None, 16, 16, 512)  2048        add_6[0][0]                      
__________________________________________________________________________________________________
leaky_re_lu_18 (LeakyReLU)      (None, 16, 16, 512)  0           batch_normalization_18[0][0]     
__________________________________________________________________________________________________
conv2d_19 (Conv2D)              (None, 16, 16, 512)  2359296     leaky_re_lu_18[0][0]             
__________________________________________________________________________________________________
batch_normalization_19 (BatchNo (None, 16, 16, 512)  2048        conv2d_19[0][0]                  
__________________________________________________________________________________________________
leaky_re_lu_19 (LeakyReLU)      (None, 16, 16, 512)  0           batch_normalization_19[0][0]     
__________________________________________________________________________________________________
conv2d_20 (Conv2D)              (None, 16, 16, 512)  2359296     leaky_re_lu_19[0][0]             
__________________________________________________________________________________________________
add_7 (Add)                     (None, 16, 16, 512)  0           conv2d_20[0][0]                  
                                                                 add_6[0][0]                      
__________________________________________________________________________________________________
batch_normalization_20 (BatchNo (None, 16, 16, 512)  2048        add_7[0][0]                      
__________________________________________________________________________________________________
leaky_re_lu_20 (LeakyReLU)      (None, 16, 16, 512)  0           batch_normalization_20[0][0]     
__________________________________________________________________________________________________
conv2d_21 (Conv2D)              (None, 16, 16, 1)    513         leaky_re_lu_20[0][0]             
__________________________________________________________________________________________________
up_sampling2d (UpSampling2D)    (None, 256, 256, 1)  0           conv2d_21[0][0]                  
==================================================================================================
Total params: 12,727,969
Trainable params: 12,718,305
Non-trainable params: 9,664
__________________________________________________________________________________________________

history = model.fit_generator(train_gen, validation_data=valid_gen, callbacks=[learning_rate,checkpoint], epochs=8, workers=8, use_multiprocessing=True,initial_epoch=6)

Epoch 7/8
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.

Epoch 00007: accuracy improved from 0.94873 to 0.95261, saving model to 4000_checkpoint_1.hdf5
 1/78 [..............................] - ETA: 0s - loss: 0.3139 - accuracy: 0.9526 - mean_iou: 0.6516
Epoch 00007: accuracy did not improve from 0.95261
 2/78 [..............................] - ETA: 32:40 - loss: 0.3334 - accuracy: 0.9474 - mean_iou: 0.6426
Epoch 00007: accuracy did not improve from 0.95261
 3/78 [>.............................] - ETA: 42:48 - loss: 0.3425 - accuracy: 0.9447 - mean_iou: 0.6287
Epoch 00007: accuracy did not improve from 0.95261
 4/78 [>.............................] - ETA: 47:22 - loss: 0.3401 - accuracy: 0.9483 - mean_iou: 0.6380
Epoch 00007: accuracy did not improve from 0.95261
 5/78 [>.............................] - ETA: 49:39 - loss: 0.3496 - accuracy: 0.9482 - mean_iou: 0.6261
Epoch 00007: accuracy did not improve from 0.95261
 6/78 [=>............................] - ETA: 50:53 - loss: 0.3624 - accuracy: 0.9449 - mean_iou: 0.6133
Epoch 00007: accuracy did not improve from 0.95261
 7/78 [=>............................] - ETA: 51:36 - loss: 0.3655 - accuracy: 0.9452 - mean_iou: 0.6192
Epoch 00007: accuracy did not improve from 0.95261
 8/78 [==>...........................] - ETA: 51:58 - loss: 0.3652 - accuracy: 0.9454 - mean_iou: 0.6129
Epoch 00007: accuracy did not improve from 0.95261
 9/78 [==>...........................] - ETA: 52:05 - loss: 0.3627 - accuracy: 0.9457 - mean_iou: 0.6222
Epoch 00007: accuracy did not improve from 0.95261
10/78 [==>...........................] - ETA: 51:55 - loss: 0.3577 - accuracy: 0.9462 - mean_iou: 0.6306
Epoch 00007: accuracy did not improve from 0.95261
11/78 [===>..........................] - ETA: 52:13 - loss: 0.3597 - accuracy: 0.9456 - mean_iou: 0.6293
Epoch 00007: accuracy did not improve from 0.95261
12/78 [===>..........................] - ETA: 51:50 - loss: 0.3568 - accuracy: 0.9463 - mean_iou: 0.6379
Epoch 00007: accuracy did not improve from 0.95261
13/78 [====>.........................] - ETA: 51:33 - loss: 0.3572 - accuracy: 0.9452 - mean_iou: 0.6321
Epoch 00007: accuracy did not improve from 0.95261
14/78 [====>.........................] - ETA: 50:58 - loss: 0.3575 - accuracy: 0.9453 - mean_iou: 0.6230
Epoch 00007: accuracy did not improve from 0.95261
15/78 [====>.........................] - ETA: 50:22 - loss: 0.3568 - accuracy: 0.9447 - mean_iou: 0.6178
Epoch 00007: accuracy did not improve from 0.95261
16/78 [=====>........................] - ETA: 49:46 - loss: 0.3598 - accuracy: 0.9445 - mean_iou: 0.6120
Epoch 00007: accuracy did not improve from 0.95261
17/78 [=====>........................] - ETA: 49:05 - loss: 0.3630 - accuracy: 0.9436 - mean_iou: 0.6110
Epoch 00007: accuracy did not improve from 0.95261
18/78 [=====>........................] - ETA: 48:23 - loss: 0.3640 - accuracy: 0.9428 - mean_iou: 0.6123
Epoch 00007: accuracy did not improve from 0.95261
19/78 [======>.......................] - ETA: 47:42 - loss: 0.3636 - accuracy: 0.9432 - mean_iou: 0.6114
Epoch 00007: accuracy did not improve from 0.95261
20/78 [======>.......................] - ETA: 46:59 - loss: 0.3643 - accuracy: 0.9426 - mean_iou: 0.6097
Epoch 00007: accuracy did not improve from 0.95261
21/78 [=======>......................] - ETA: 46:17 - loss: 0.3628 - accuracy: 0.9433 - mean_iou: 0.6161
Epoch 00007: accuracy did not improve from 0.95261
22/78 [=======>......................] - ETA: 45:33 - loss: 0.3617 - accuracy: 0.9437 - mean_iou: 0.6171
Epoch 00007: accuracy did not improve from 0.95261
23/78 [=======>......................] - ETA: 44:49 - loss: 0.3612 - accuracy: 0.9439 - mean_iou: 0.6169
Epoch 00007: accuracy did not improve from 0.95261
24/78 [========>.....................] - ETA: 44:05 - loss: 0.3635 - accuracy: 0.9441 - mean_iou: 0.6142
Epoch 00007: accuracy did not improve from 0.95261
25/78 [========>.....................] - ETA: 43:20 - loss: 0.3656 - accuracy: 0.9437 - mean_iou: 0.6108
Epoch 00007: accuracy did not improve from 0.95261
26/78 [=========>....................] - ETA: 42:35 - loss: 0.3662 - accuracy: 0.9434 - mean_iou: 0.6092
Epoch 00007: accuracy did not improve from 0.95261
27/78 [=========>....................] - ETA: 41:49 - loss: 0.3644 - accuracy: 0.9438 - mean_iou: 0.6150
Epoch 00007: accuracy did not improve from 0.95261
28/78 [=========>....................] - ETA: 41:02 - loss: 0.3651 - accuracy: 0.9442 - mean_iou: 0.6139
Epoch 00007: accuracy did not improve from 0.95261
29/78 [==========>...................] - ETA: 40:16 - loss: 0.3641 - accuracy: 0.9444 - mean_iou: 0.6135
Epoch 00007: accuracy did not improve from 0.95261
30/78 [==========>...................] - ETA: 39:28 - loss: 0.3644 - accuracy: 0.9443 - mean_iou: 0.6125
Epoch 00007: accuracy did not improve from 0.95261
31/78 [==========>...................] - ETA: 38:41 - loss: 0.3647 - accuracy: 0.9442 - mean_iou: 0.6131
Epoch 00007: accuracy did not improve from 0.95261
32/78 [===========>..................] - ETA: 37:53 - loss: 0.3652 - accuracy: 0.9442 - mean_iou: 0.6113
Epoch 00007: accuracy did not improve from 0.95261
33/78 [===========>..................] - ETA: 37:05 - loss: 0.3655 - accuracy: 0.9438 - mean_iou: 0.6134
Epoch 00007: accuracy did not improve from 0.95261
34/78 [============>.................] - ETA: 36:17 - loss: 0.3654 - accuracy: 0.9433 - mean_iou: 0.6123
Epoch 00007: accuracy did not improve from 0.95261
35/78 [============>.................] - ETA: 35:29 - loss: 0.3656 - accuracy: 0.9433 - mean_iou: 0.6099
Epoch 00007: accuracy did not improve from 0.95261
36/78 [============>.................] - ETA: 34:41 - loss: 0.3669 - accuracy: 0.9428 - mean_iou: 0.6060
Epoch 00007: accuracy did not improve from 0.95261
37/78 [=============>................] - ETA: 33:53 - loss: 0.3677 - accuracy: 0.9427 - mean_iou: 0.6058
Epoch 00007: accuracy did not improve from 0.95261
38/78 [=============>................] - ETA: 33:04 - loss: 0.3673 - accuracy: 0.9426 - mean_iou: 0.6065
Epoch 00007: accuracy did not improve from 0.95261
39/78 [==============>...............] - ETA: 32:16 - loss: 0.3672 - accuracy: 0.9427 - mean_iou: 0.6050
Epoch 00007: accuracy did not improve from 0.95261
40/78 [==============>...............] - ETA: 31:27 - loss: 0.3679 - accuracy: 0.9426 - mean_iou: 0.6040
Epoch 00007: accuracy did not improve from 0.95261
41/78 [==============>...............] - ETA: 30:38 - loss: 0.3676 - accuracy: 0.9424 - mean_iou: 0.6026
Epoch 00007: accuracy did not improve from 0.95261
42/78 [===============>..............] - ETA: 29:49 - loss: 0.3669 - accuracy: 0.9423 - mean_iou: 0.6026
Epoch 00007: accuracy did not improve from 0.95261
43/78 [===============>..............] - ETA: 29:00 - loss: 0.3671 - accuracy: 0.9423 - mean_iou: 0.6024
Epoch 00007: accuracy did not improve from 0.95261
44/78 [===============>..............] - ETA: 28:11 - loss: 0.3662 - accuracy: 0.9424 - mean_iou: 0.6035
Epoch 00007: accuracy did not improve from 0.95261
45/78 [================>.............] - ETA: 27:22 - loss: 0.3664 - accuracy: 0.9422 - mean_iou: 0.6026
Epoch 00007: accuracy did not improve from 0.95261
46/78 [================>.............] - ETA: 26:33 - loss: 0.3664 - accuracy: 0.9422 - mean_iou: 0.6032
Epoch 00007: accuracy did not improve from 0.95261
47/78 [=================>............] - ETA: 25:44 - loss: 0.3657 - accuracy: 0.9423 - mean_iou: 0.6060
Epoch 00007: accuracy did not improve from 0.95261
48/78 [=================>............] - ETA: 24:55 - loss: 0.3653 - accuracy: 0.9425 - mean_iou: 0.6057
Epoch 00007: accuracy did not improve from 0.95261
49/78 [=================>............] - ETA: 24:05 - loss: 0.3646 - accuracy: 0.9427 - mean_iou: 0.6068
Epoch 00007: accuracy did not improve from 0.95261
50/78 [==================>...........] - ETA: 23:16 - loss: 0.3632 - accuracy: 0.9431 - mean_iou: 0.6086
Epoch 00007: accuracy did not improve from 0.95261
51/78 [==================>...........] - ETA: 22:26 - loss: 0.3625 - accuracy: 0.9432 - mean_iou: 0.6087
Epoch 00007: accuracy did not improve from 0.95261
52/78 [===================>..........] - ETA: 21:37 - loss: 0.3621 - accuracy: 0.9432 - mean_iou: 0.6079
Epoch 00007: accuracy did not improve from 0.95261
53/78 [===================>..........] - ETA: 20:47 - loss: 0.3623 - accuracy: 0.9432 - mean_iou: 0.6064
Epoch 00007: accuracy did not improve from 0.95261
54/78 [===================>..........] - ETA: 19:58 - loss: 0.3624 - accuracy: 0.9431 - mean_iou: 0.6056
Epoch 00007: accuracy did not improve from 0.95261
55/78 [====================>.........] - ETA: 19:08 - loss: 0.3625 - accuracy: 0.9432 - mean_iou: 0.6051
Epoch 00007: accuracy did not improve from 0.95261
56/78 [====================>.........] - ETA: 18:18 - loss: 0.3650 - accuracy: 0.9426 - mean_iou: 0.6042
Epoch 00007: accuracy did not improve from 0.95261
57/78 [====================>.........] - ETA: 17:29 - loss: 0.3649 - accuracy: 0.9426 - mean_iou: 0.6053
Epoch 00007: accuracy did not improve from 0.95261
58/78 [=====================>........] - ETA: 16:39 - loss: 0.3651 - accuracy: 0.9426 - mean_iou: 0.6047
Epoch 00007: accuracy did not improve from 0.95261
59/78 [=====================>........] - ETA: 15:49 - loss: 0.3665 - accuracy: 0.9425 - mean_iou: 0.6039
Epoch 00007: accuracy did not improve from 0.95261
60/78 [======================>.......] - ETA: 14:59 - loss: 0.3669 - accuracy: 0.9424 - mean_iou: 0.6022
Epoch 00007: accuracy did not improve from 0.95261
61/78 [======================>.......] - ETA: 14:10 - loss: 0.3673 - accuracy: 0.9424 - mean_iou: 0.6006
Epoch 00007: accuracy did not improve from 0.95261
62/78 [======================>.......] - ETA: 13:20 - loss: 0.3675 - accuracy: 0.9423 - mean_iou: 0.5991
Epoch 00007: accuracy did not improve from 0.95261
63/78 [=======================>......] - ETA: 12:30 - loss: 0.3678 - accuracy: 0.9422 - mean_iou: 0.5998
Epoch 00007: accuracy did not improve from 0.95261
64/78 [=======================>......] - ETA: 11:40 - loss: 0.3672 - accuracy: 0.9422 - mean_iou: 0.6013
Epoch 00007: accuracy did not improve from 0.95261
65/78 [========================>.....] - ETA: 10:51 - loss: 0.3663 - accuracy: 0.9422 - mean_iou: 0.6029
Epoch 00007: accuracy did not improve from 0.95261
66/78 [========================>.....] - ETA: 10:01 - loss: 0.3658 - accuracy: 0.9424 - mean_iou: 0.6038
Epoch 00007: accuracy did not improve from 0.95261
67/78 [========================>.....] - ETA: 9:11 - loss: 0.3655 - accuracy: 0.9425 - mean_iou: 0.6052 
Epoch 00007: accuracy did not improve from 0.95261
68/78 [=========================>....] - ETA: 8:20 - loss: 0.3659 - accuracy: 0.9424 - mean_iou: 0.6044
Epoch 00007: accuracy did not improve from 0.95261
69/78 [=========================>....] - ETA: 7:29 - loss: 0.3662 - accuracy: 0.9424 - mean_iou: 0.6036
Epoch 00007: accuracy did not improve from 0.95261
70/78 [=========================>....] - ETA: 6:39 - loss: 0.3661 - accuracy: 0.9425 - mean_iou: 0.6047
Epoch 00007: accuracy did not improve from 0.95261
71/78 [==========================>...] - ETA: 5:49 - loss: 0.3656 - accuracy: 0.9427 - mean_iou: 0.6064
Epoch 00007: accuracy did not improve from 0.95261
72/78 [==========================>...] - ETA: 4:59 - loss: 0.3658 - accuracy: 0.9425 - mean_iou: 0.6066
Epoch 00007: accuracy did not improve from 0.95261
73/78 [===========================>..] - ETA: 4:08 - loss: 0.3664 - accuracy: 0.9425 - mean_iou: 0.6051
Epoch 00007: accuracy did not improve from 0.95261
74/78 [===========================>..] - ETA: 3:19 - loss: 0.3660 - accuracy: 0.9426 - mean_iou: 0.6057
Epoch 00007: accuracy did not improve from 0.95261
75/78 [===========================>..] - ETA: 2:29 - loss: 0.3662 - accuracy: 0.9427 - mean_iou: 0.6050
Epoch 00007: accuracy did not improve from 0.95261
76/78 [============================>.] - ETA: 1:39 - loss: 0.3667 - accuracy: 0.9428 - mean_iou: 0.6044
Epoch 00007: accuracy did not improve from 0.95261
77/78 [============================>.] - ETA: 49s - loss: 0.3665 - accuracy: 0.9428 - mean_iou: 0.6046 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.

Epoch 00007: accuracy did not improve from 0.95261
78/78 [==============================] - ETA: 0s - loss: 0.3665 - accuracy: 0.9427 - mean_iou: 0.6042 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
78/78 [==============================] - 4159s 53s/step - loss: 0.3665 - accuracy: 0.9427 - mean_iou: 0.6042 - val_loss: 0.3991 - val_accuracy: 0.9474 - val_mean_iou: 0.5798 - lr: 8.6448e-04
Epoch 8/8
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.

Epoch 00008: accuracy did not improve from 0.95261
 1/78 [..............................] - ETA: 0s - loss: 0.3226 - accuracy: 0.9474 - mean_iou: 0.7384
Epoch 00008: accuracy did not improve from 0.95261
 2/78 [..............................] - ETA: 32:11 - loss: 0.3244 - accuracy: 0.9507 - mean_iou: 0.7047
Epoch 00008: accuracy did not improve from 0.95261
 3/78 [>.............................] - ETA: 42:13 - loss: 0.3402 - accuracy: 0.9468 - mean_iou: 0.6832
Epoch 00008: accuracy did not improve from 0.95261
 4/78 [>.............................] - ETA: 46:52 - loss: 0.3572 - accuracy: 0.9452 - mean_iou: 0.6669
Epoch 00008: accuracy did not improve from 0.95261
 5/78 [>.............................] - ETA: 49:18 - loss: 0.3509 - accuracy: 0.9465 - mean_iou: 0.6540
Epoch 00008: accuracy did not improve from 0.95261
 6/78 [=>............................] - ETA: 50:38 - loss: 0.3512 - accuracy: 0.9462 - mean_iou: 0.6470
Epoch 00008: accuracy did not improve from 0.95261
 7/78 [=>............................] - ETA: 51:20 - loss: 0.3490 - accuracy: 0.9458 - mean_iou: 0.6486
Epoch 00008: accuracy did not improve from 0.95261
 8/78 [==>...........................] - ETA: 51:37 - loss: 0.3478 - accuracy: 0.9461 - mean_iou: 0.6522
Epoch 00008: accuracy did not improve from 0.95261
 9/78 [==>...........................] - ETA: 51:42 - loss: 0.3521 - accuracy: 0.9438 - mean_iou: 0.6419
Epoch 00008: accuracy did not improve from 0.95261
10/78 [==>...........................] - ETA: 51:35 - loss: 0.3531 - accuracy: 0.9437 - mean_iou: 0.6397
Epoch 00008: accuracy did not improve from 0.95261
11/78 [===>..........................] - ETA: 51:19 - loss: 0.3518 - accuracy: 0.9438 - mean_iou: 0.6468
Epoch 00008: accuracy did not improve from 0.95261
12/78 [===>..........................] - ETA: 51:01 - loss: 0.3575 - accuracy: 0.9430 - mean_iou: 0.6415
Epoch 00008: accuracy did not improve from 0.95261
13/78 [====>.........................] - ETA: 50:33 - loss: 0.3556 - accuracy: 0.9443 - mean_iou: 0.6375
Epoch 00008: accuracy did not improve from 0.95261
14/78 [====>.........................] - ETA: 50:03 - loss: 0.3568 - accuracy: 0.9453 - mean_iou: 0.6431
Epoch 00008: accuracy did not improve from 0.95261
15/78 [====>.........................] - ETA: 49:30 - loss: 0.3551 - accuracy: 0.9460 - mean_iou: 0.6434
Epoch 00008: accuracy did not improve from 0.95261
16/78 [=====>........................] - ETA: 48:56 - loss: 0.3562 - accuracy: 0.9457 - mean_iou: 0.6418
Epoch 00008: accuracy did not improve from 0.95261
17/78 [=====>........................] - ETA: 48:19 - loss: 0.3571 - accuracy: 0.9462 - mean_iou: 0.6422
Epoch 00008: accuracy did not improve from 0.95261
18/78 [=====>........................] - ETA: 47:42 - loss: 0.3602 - accuracy: 0.9454 - mean_iou: 0.6371
Epoch 00008: accuracy did not improve from 0.95261
19/78 [======>.......................] - ETA: 47:03 - loss: 0.3612 - accuracy: 0.9454 - mean_iou: 0.6280
Epoch 00008: accuracy did not improve from 0.95261
20/78 [======>.......................] - ETA: 46:23 - loss: 0.3609 - accuracy: 0.9454 - mean_iou: 0.6299
Epoch 00008: accuracy did not improve from 0.95261
21/78 [=======>......................] - ETA: 45:42 - loss: 0.3610 - accuracy: 0.9454 - mean_iou: 0.6303
Epoch 00008: accuracy did not improve from 0.95261
22/78 [=======>......................] - ETA: 44:59 - loss: 0.3629 - accuracy: 0.9441 - mean_iou: 0.6273
Epoch 00008: accuracy did not improve from 0.95261
23/78 [=======>......................] - ETA: 44:16 - loss: 0.3607 - accuracy: 0.9445 - mean_iou: 0.6292
Epoch 00008: accuracy did not improve from 0.95261
24/78 [========>.....................] - ETA: 43:32 - loss: 0.3596 - accuracy: 0.9444 - mean_iou: 0.6279
Epoch 00008: accuracy did not improve from 0.95261
25/78 [========>.....................] - ETA: 42:47 - loss: 0.3587 - accuracy: 0.9444 - mean_iou: 0.6257
Epoch 00008: accuracy did not improve from 0.95261
26/78 [=========>....................] - ETA: 42:02 - loss: 0.3594 - accuracy: 0.9439 - mean_iou: 0.6232
Epoch 00008: accuracy did not improve from 0.95261
27/78 [=========>....................] - ETA: 41:17 - loss: 0.3606 - accuracy: 0.9440 - mean_iou: 0.6245
Epoch 00008: accuracy did not improve from 0.95261
28/78 [=========>....................] - ETA: 40:32 - loss: 0.3599 - accuracy: 0.9442 - mean_iou: 0.6264
Epoch 00008: accuracy did not improve from 0.95261
29/78 [==========>...................] - ETA: 39:46 - loss: 0.3622 - accuracy: 0.9437 - mean_iou: 0.6196
Epoch 00008: accuracy did not improve from 0.95261
30/78 [==========>...................] - ETA: 39:00 - loss: 0.3636 - accuracy: 0.9433 - mean_iou: 0.6193
Epoch 00008: accuracy did not improve from 0.95261
31/78 [==========>...................] - ETA: 38:14 - loss: 0.3637 - accuracy: 0.9431 - mean_iou: 0.6187
Epoch 00008: accuracy did not improve from 0.95261
32/78 [===========>..................] - ETA: 37:27 - loss: 0.3640 - accuracy: 0.9430 - mean_iou: 0.6200
Epoch 00008: accuracy did not improve from 0.95261
33/78 [===========>..................] - ETA: 36:41 - loss: 0.3633 - accuracy: 0.9432 - mean_iou: 0.6185
Epoch 00008: accuracy did not improve from 0.95261
34/78 [============>.................] - ETA: 35:54 - loss: 0.3624 - accuracy: 0.9435 - mean_iou: 0.6176
Epoch 00008: accuracy did not improve from 0.95261
35/78 [============>.................] - ETA: 35:06 - loss: 0.3623 - accuracy: 0.9435 - mean_iou: 0.6169
Epoch 00008: accuracy did not improve from 0.95261
36/78 [============>.................] - ETA: 34:19 - loss: 0.3626 - accuracy: 0.9434 - mean_iou: 0.6157
Epoch 00008: accuracy did not improve from 0.95261
37/78 [=============>................] - ETA: 33:31 - loss: 0.3630 - accuracy: 0.9433 - mean_iou: 0.6158
Epoch 00008: accuracy did not improve from 0.95261
38/78 [=============>................] - ETA: 32:44 - loss: 0.3633 - accuracy: 0.9434 - mean_iou: 0.6118
Epoch 00008: accuracy did not improve from 0.95261
39/78 [==============>...............] - ETA: 31:55 - loss: 0.3631 - accuracy: 0.9433 - mean_iou: 0.6135
Epoch 00008: accuracy did not improve from 0.95261
40/78 [==============>...............] - ETA: 31:08 - loss: 0.3619 - accuracy: 0.9436 - mean_iou: 0.6155
Epoch 00008: accuracy did not improve from 0.95261
41/78 [==============>...............] - ETA: 30:20 - loss: 0.3610 - accuracy: 0.9438 - mean_iou: 0.6162
Epoch 00008: accuracy did not improve from 0.95261
42/78 [===============>..............] - ETA: 29:32 - loss: 0.3618 - accuracy: 0.9436 - mean_iou: 0.6169
Epoch 00008: accuracy did not improve from 0.95261
43/78 [===============>..............] - ETA: 28:44 - loss: 0.3620 - accuracy: 0.9436 - mean_iou: 0.6167
Epoch 00008: accuracy did not improve from 0.95261
44/78 [===============>..............] - ETA: 27:56 - loss: 0.3614 - accuracy: 0.9438 - mean_iou: 0.6174
Epoch 00008: accuracy did not improve from 0.95261
45/78 [================>.............] - ETA: 27:08 - loss: 0.3610 - accuracy: 0.9440 - mean_iou: 0.6180
Epoch 00008: accuracy did not improve from 0.95261
46/78 [================>.............] - ETA: 26:19 - loss: 0.3604 - accuracy: 0.9443 - mean_iou: 0.6179
Epoch 00008: accuracy did not improve from 0.95261
47/78 [=================>............] - ETA: 25:31 - loss: 0.3606 - accuracy: 0.9443 - mean_iou: 0.6157
Epoch 00008: accuracy did not improve from 0.95261
48/78 [=================>............] - ETA: 24:43 - loss: 0.3607 - accuracy: 0.9441 - mean_iou: 0.6161
Epoch 00008: accuracy did not improve from 0.95261
49/78 [=================>............] - ETA: 23:54 - loss: 0.3610 - accuracy: 0.9441 - mean_iou: 0.6156
Epoch 00008: accuracy did not improve from 0.95261
50/78 [==================>...........] - ETA: 23:05 - loss: 0.3615 - accuracy: 0.9440 - mean_iou: 0.6163
Epoch 00008: accuracy did not improve from 0.95261
51/78 [==================>...........] - ETA: 22:17 - loss: 0.3610 - accuracy: 0.9441 - mean_iou: 0.6155
Epoch 00008: accuracy did not improve from 0.95261
52/78 [===================>..........] - ETA: 21:28 - loss: 0.3613 - accuracy: 0.9440 - mean_iou: 0.6146
Epoch 00008: accuracy did not improve from 0.95261
53/78 [===================>..........] - ETA: 20:39 - loss: 0.3619 - accuracy: 0.9437 - mean_iou: 0.6125
Epoch 00008: accuracy did not improve from 0.95261
54/78 [===================>..........] - ETA: 19:50 - loss: 0.3608 - accuracy: 0.9440 - mean_iou: 0.6131
Epoch 00008: accuracy did not improve from 0.95261
55/78 [====================>.........] - ETA: 19:01 - loss: 0.3607 - accuracy: 0.9439 - mean_iou: 0.6112
Epoch 00008: accuracy did not improve from 0.95261
56/78 [====================>.........] - ETA: 18:11 - loss: 0.3611 - accuracy: 0.9434 - mean_iou: 0.6101
Epoch 00008: accuracy did not improve from 0.95261
57/78 [====================>.........] - ETA: 17:22 - loss: 0.3615 - accuracy: 0.9433 - mean_iou: 0.6093
Epoch 00008: accuracy did not improve from 0.95261
58/78 [=====================>........] - ETA: 16:33 - loss: 0.3615 - accuracy: 0.9432 - mean_iou: 0.6079
Epoch 00008: accuracy did not improve from 0.95261
59/78 [=====================>........] - ETA: 15:43 - loss: 0.3616 - accuracy: 0.9430 - mean_iou: 0.6079
Epoch 00008: accuracy did not improve from 0.95261
60/78 [======================>.......] - ETA: 14:54 - loss: 0.3610 - accuracy: 0.9430 - mean_iou: 0.6075
Epoch 00008: accuracy did not improve from 0.95261
61/78 [======================>.......] - ETA: 14:05 - loss: 0.3624 - accuracy: 0.9428 - mean_iou: 0.6058
Epoch 00008: accuracy did not improve from 0.95261
62/78 [======================>.......] - ETA: 13:15 - loss: 0.3624 - accuracy: 0.9427 - mean_iou: 0.6066
Epoch 00008: accuracy did not improve from 0.95261
63/78 [=======================>......] - ETA: 12:25 - loss: 0.3626 - accuracy: 0.9429 - mean_iou: 0.6065
Epoch 00008: accuracy did not improve from 0.95261
64/78 [=======================>......] - ETA: 11:36 - loss: 0.3620 - accuracy: 0.9430 - mean_iou: 0.6066
Epoch 00008: accuracy did not improve from 0.95261
65/78 [========================>.....] - ETA: 10:46 - loss: 0.3611 - accuracy: 0.9432 - mean_iou: 0.6072
Epoch 00008: accuracy did not improve from 0.95261
66/78 [========================>.....] - ETA: 9:57 - loss: 0.3610 - accuracy: 0.9432 - mean_iou: 0.6074 
Epoch 00008: accuracy did not improve from 0.95261
67/78 [========================>.....] - ETA: 9:07 - loss: 0.3606 - accuracy: 0.9435 - mean_iou: 0.6090
Epoch 00008: accuracy did not improve from 0.95261
68/78 [=========================>....] - ETA: 8:17 - loss: 0.3599 - accuracy: 0.9437 - mean_iou: 0.6101
Epoch 00008: accuracy did not improve from 0.95261
69/78 [=========================>....] - ETA: 7:27 - loss: 0.3589 - accuracy: 0.9440 - mean_iou: 0.6106
Epoch 00008: accuracy did not improve from 0.95261
70/78 [=========================>....] - ETA: 6:37 - loss: 0.3594 - accuracy: 0.9441 - mean_iou: 0.6110
Epoch 00008: accuracy did not improve from 0.95261
71/78 [==========================>...] - ETA: 5:47 - loss: 0.3594 - accuracy: 0.9440 - mean_iou: 0.6112
Epoch 00008: accuracy did not improve from 0.95261
72/78 [==========================>...] - ETA: 4:57 - loss: 0.3600 - accuracy: 0.9438 - mean_iou: 0.6100
Epoch 00008: accuracy did not improve from 0.95261
73/78 [===========================>..] - ETA: 4:07 - loss: 0.3595 - accuracy: 0.9439 - mean_iou: 0.6108
Epoch 00008: accuracy did not improve from 0.95261
74/78 [===========================>..] - ETA: 3:18 - loss: 0.3594 - accuracy: 0.9439 - mean_iou: 0.6107
Epoch 00008: accuracy did not improve from 0.95261
75/78 [===========================>..] - ETA: 2:28 - loss: 0.3593 - accuracy: 0.9439 - mean_iou: 0.6113
Epoch 00008: accuracy did not improve from 0.95261
76/78 [============================>.] - ETA: 1:38 - loss: 0.3607 - accuracy: 0.9437 - mean_iou: 0.6097
Epoch 00008: accuracy did not improve from 0.95261
77/78 [============================>.] - ETA: 49s - loss: 0.3599 - accuracy: 0.9438 - mean_iou: 0.6102 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.

Epoch 00008: accuracy did not improve from 0.95261
78/78 [==============================] - ETA: 0s - loss: 0.3597 - accuracy: 0.9438 - mean_iou: 0.6100 WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
WARNING:tensorflow:multiprocessing can interact badly with TensorFlow, causing nondeterministic deadlocks. For high performance data pipelines tf.data is recommended.
78/78 [==============================] - 4147s 53s/step - loss: 0.3597 - accuracy: 0.9438 - mean_iou: 0.6100 - val_loss: 0.3901 - val_accuracy: 0.9344 - val_mean_iou: 0.4983 - lr: 8.1871e-04

model_json = model.to_json()
with open("model4000_8.json", "w") as json_file:
    json_file.write(model_json)

model.save_weights("model4000_8.h5")

plt.figure(figsize=(12,4))
plt.subplot(131)
plt.plot(history.epoch, history.history["loss"], label="Train loss")
plt.plot(history.epoch, history.history["val_loss"], label="Valid loss")
plt.legend()
plt.subplot(132)
plt.plot(history.epoch, history.history["accuracy"], label="Train accuracy")
plt.plot(history.epoch, history.history["val_accuracy"], label="Valid accuracy")
plt.legend()
plt.subplot(133)
plt.plot(history.epoch, history.history["mean_iou"], label="Train iou")
plt.plot(history.epoch, history.history["val_mean_iou"], label="Valid iou")
plt.legend()
plt.show()

Above training we have done for additional 2 epochs.

Prediction on validation set¶

for imgs, msks in valid_gen:
    # predict batch of images
    preds = model.predict(imgs)
    # create figure
    f, axarr = plt.subplots(4, 8, figsize=(20,15))
    axarr = axarr.ravel()
    axidx = 0
    # loop through batch
    for img, msk, pred in zip(imgs, msks, preds):
        # plot image
        axarr[axidx].imshow(img[:, :, 0])
        # threshold true mask
        comp = msk[:, :, 0] > 0.5
        # apply connected components
        comp = measure.label(comp)
        # apply bounding boxes
        predictionString = ''
        for region in measure.regionprops(comp):
            # retrieve x, y, height and width
            y, x, y2, x2 = region.bbox
            height = y2 - y
            width = x2 - x
            axarr[axidx].add_patch(patches.Rectangle((x,y),width,height,linewidth=2,edgecolor='b',facecolor='none'))
        # threshold predicted mask
        comp = pred[:, :, 0] > 0.5
        # apply connected components
        comp = measure.label(comp)
        # apply bounding boxes
        predictionString = ''
        for region in measure.regionprops(comp):
            # retrieve x, y, height and width
            y, x, y2, x2 = region.bbox
            height = y2 - y
            width = x2 - x
            conf = np.mean(pred[y:y+height, x:x+width])
            if conf>0.3:
              axarr[axidx].add_patch(patches.Rectangle((x,y),width,height,linewidth=2,edgecolor='r',facecolor='none'))
        axidx += 1
    plt.show()
    # only plot one batch
    break

Prediction on test set¶

folder = 'stage_2_test_images'
test_filenames = os.listdir(folder)
# create test generator with predict flag set to True
test_gen = generator(folder, test_filenames, None, batch_size=2, image_size=256, shuffle=False, predict=True)

print('n test samples:', len(test_filenames))
f, axarr = plt.subplots(2, 5, figsize=(20,10))
axarr = axarr.ravel()
axidx = 0
# create submission dictionary
submission_dict = {}
# loop through testset
for imgs, filenames in test_gen:
    # predict batch of images
    preds = model.predict(imgs)
    #axarr[axidx].imshow(resize(img[:, :, 0],(1024, 1024),mode='reflect'))
    axarr[axidx].imshow(img[:, :, 0])
    # loop through batch
    for pred, filename in zip(preds, filenames):
        # resize predicted mask
        #pred = resize(pred, (1024, 1024), mode='reflect')
        # threshold predicted mask
        comp = pred[:, :, 0] > 0.5
        # apply connected components
        comp = measure.label(comp)
        # apply bounding boxes
        predictionString = ''
        for region in measure.regionprops(comp):
            # retrieve x, y, height and width
            y, x, y2, x2 = region.bbox
            height = y2 - y
            width = x2 - x
            # proxy for confidence score
            conf = np.mean(pred[y:y+height, x:x+width])
            # add to predictionString
            if conf>0.8:
              predictionString += str(conf) + ' ' + str(x) + ' ' + str(y) + ' ' + str(width) + ' ' + str(height) + ' '
              axarr[axidx].add_patch(patches.Rectangle((x,y),width,height,linewidth=2,edgecolor='r',facecolor='none'))
        # add filename and predictionString to dictionary
        filename = filename.split('.')[0]
        submission_dict[filename] = predictionString
    #print("--------------------------------------------------------------------------")
    #print(axidx)
    #print(predictionString)
    axidx += 1
    if axidx >= 10: #len(test_filenames)
        break
plt.show()

n test samples: 3000

DenseNet121¶

DenseNet falls in the category of classic networks. DenseNet is quite similar to ResNet with some fundamental differences. ResNet uses an additive method (+) that merges the previous layer (identity) with the future layer, whereas DenseNet concatenates (.) the output of the previous layer with the future layer.

from keras.layers import *
from keras.applications.densenet import DenseNet121
from keras.models import Model
base_model=DenseNet121(include_top=False, weights= None,  input_shape=(256,256,1))
base_model.trainable = True
print(base_model.input)
x = Dense(1, activation='sigmoid')(base_model.output)
x = UpSampling2D(32)(x)
print(x)
transfer_model = Model(base_model.input, x) 
print(x)
transfer_model.compile(optimizer = 'adam', 
              loss=iou_bce_loss,
              metrics=['accuracy'])

Tensor("input_4:0", shape=(None, 256, 256, 1), dtype=float32)
Tensor("up_sampling2d_3/ResizeNearestNeighbor:0", shape=(None, 256, 256, 1), dtype=float32)
Tensor("up_sampling2d_3/ResizeNearestNeighbor:0", shape=(None, 256, 256, 1), dtype=float32)

history1 = transfer_model.fit_generator(train_gen, validation_data=valid_gen, callbacks=[learning_rate], epochs=5)

Epoch 1/5
75/75 [==============================] - 2233s 30s/step - loss: 0.5179 - accuracy: 0.8895 - val_loss: 0.9258 - val_accuracy: 0.4323
Epoch 2/5
75/75 [==============================] - 2156s 29s/step - loss: 0.4605 - accuracy: 0.9160 - val_loss: 0.7090 - val_accuracy: 0.7760
Epoch 3/5
75/75 [==============================] - 2200s 29s/step - loss: 0.4484 - accuracy: 0.9195 - val_loss: 0.7891 - val_accuracy: 0.6217
Epoch 4/5
75/75 [==============================] - 2193s 29s/step - loss: 0.4459 - accuracy: 0.9200 - val_loss: 0.6942 - val_accuracy: 0.7713
Epoch 5/5
75/75 [==============================] - 2266s 30s/step - loss: 0.4376 - accuracy: 0.9207 - val_loss: 0.4804 - val_accuracy: 0.9167

plt.figure(figsize=(12,4))
plt.subplot(131)
plt.plot(history1.epoch, history1.history["loss"], label="Train loss")
plt.plot(history1.epoch, history1.history["val_loss"], label="Valid loss")
plt.legend()
plt.subplot(132)
plt.plot(history1.epoch, history1.history["accuracy"], label="Train accuracy")
plt.plot(history1.epoch, history1.history["val_accuracy"], label="Valid accuracy")
plt.legend()
plt.show()

transfer_model.summary()

Model: "model_2"
__________________________________________________________________________________________________
Layer (type)                    Output Shape         Param #     Connected to                     
==================================================================================================
input_4 (InputLayer)            (None, 256, 256, 1)  0                                            
__________________________________________________________________________________________________
zero_padding2d_5 (ZeroPadding2D (None, 262, 262, 1)  0           input_4[0][0]                    
__________________________________________________________________________________________________
conv1/conv (Conv2D)             (None, 128, 128, 64) 3136        zero_padding2d_5[0][0]           
__________________________________________________________________________________________________
conv1/bn (BatchNormalization)   (None, 128, 128, 64) 256         conv1/conv[0][0]                 
__________________________________________________________________________________________________
conv1/relu (Activation)         (None, 128, 128, 64) 0           conv1/bn[0][0]                   
__________________________________________________________________________________________________
zero_padding2d_6 (ZeroPadding2D (None, 130, 130, 64) 0           conv1/relu[0][0]                 
__________________________________________________________________________________________________
pool1 (MaxPooling2D)            (None, 64, 64, 64)   0           zero_padding2d_6[0][0]           
__________________________________________________________________________________________________
conv2_block1_0_bn (BatchNormali (None, 64, 64, 64)   256         pool1[0][0]                      
__________________________________________________________________________________________________
conv2_block1_0_relu (Activation (None, 64, 64, 64)   0           conv2_block1_0_bn[0][0]          
__________________________________________________________________________________________________
conv2_block1_1_conv (Conv2D)    (None, 64, 64, 128)  8192        conv2_block1_0_relu[0][0]        
__________________________________________________________________________________________________
conv2_block1_1_bn (BatchNormali (None, 64, 64, 128)  512         conv2_block1_1_conv[0][0]        
__________________________________________________________________________________________________
conv2_block1_1_relu (Activation (None, 64, 64, 128)  0           conv2_block1_1_bn[0][0]          
__________________________________________________________________________________________________
conv2_block1_2_conv (Conv2D)    (None, 64, 64, 32)   36864       conv2_block1_1_relu[0][0]        
__________________________________________________________________________________________________
conv2_block1_concat (Concatenat (None, 64, 64, 96)   0           pool1[0][0]                      
                                                                 conv2_block1_2_conv[0][0]        
__________________________________________________________________________________________________
conv2_block2_0_bn (BatchNormali (None, 64, 64, 96)   384         conv2_block1_concat[0][0]        
__________________________________________________________________________________________________
conv2_block2_0_relu (Activation (None, 64, 64, 96)   0           conv2_block2_0_bn[0][0]          
__________________________________________________________________________________________________
conv2_block2_1_conv (Conv2D)    (None, 64, 64, 128)  12288       conv2_block2_0_relu[0][0]        
__________________________________________________________________________________________________
conv2_block2_1_bn (BatchNormali (None, 64, 64, 128)  512         conv2_block2_1_conv[0][0]        
__________________________________________________________________________________________________
conv2_block2_1_relu (Activation (None, 64, 64, 128)  0           conv2_block2_1_bn[0][0]          
__________________________________________________________________________________________________
conv2_block2_2_conv (Conv2D)    (None, 64, 64, 32)   36864       conv2_block2_1_relu[0][0]        
__________________________________________________________________________________________________
conv2_block2_concat (Concatenat (None, 64, 64, 128)  0           conv2_block1_concat[0][0]        
                                                                 conv2_block2_2_conv[0][0]        
__________________________________________________________________________________________________
conv2_block3_0_bn (BatchNormali (None, 64, 64, 128)  512         conv2_block2_concat[0][0]        
__________________________________________________________________________________________________
conv2_block3_0_relu (Activation (None, 64, 64, 128)  0           conv2_block3_0_bn[0][0]          
__________________________________________________________________________________________________
conv2_block3_1_conv (Conv2D)    (None, 64, 64, 128)  16384       conv2_block3_0_relu[0][0]        
__________________________________________________________________________________________________
conv2_block3_1_bn (BatchNormali (None, 64, 64, 128)  512         conv2_block3_1_conv[0][0]        
__________________________________________________________________________________________________
conv2_block3_1_relu (Activation (None, 64, 64, 128)  0           conv2_block3_1_bn[0][0]          
__________________________________________________________________________________________________
conv2_block3_2_conv (Conv2D)    (None, 64, 64, 32)   36864       conv2_block3_1_relu[0][0]        
__________________________________________________________________________________________________
conv2_block3_concat (Concatenat (None, 64, 64, 160)  0           conv2_block2_concat[0][0]        
                                                                 conv2_block3_2_conv[0][0]        
__________________________________________________________________________________________________
conv2_block4_0_bn (BatchNormali (None, 64, 64, 160)  640         conv2_block3_concat[0][0]        
__________________________________________________________________________________________________
conv2_block4_0_relu (Activation (None, 64, 64, 160)  0           conv2_block4_0_bn[0][0]          
__________________________________________________________________________________________________
conv2_block4_1_conv (Conv2D)    (None, 64, 64, 128)  20480       conv2_block4_0_relu[0][0]        
__________________________________________________________________________________________________
conv2_block4_1_bn (BatchNormali (None, 64, 64, 128)  512         conv2_block4_1_conv[0][0]        
__________________________________________________________________________________________________
conv2_block4_1_relu (Activation (None, 64, 64, 128)  0           conv2_block4_1_bn[0][0]          
__________________________________________________________________________________________________
conv2_block4_2_conv (Conv2D)    (None, 64, 64, 32)   36864       conv2_block4_1_relu[0][0]        
__________________________________________________________________________________________________
conv2_block4_concat (Concatenat (None, 64, 64, 192)  0           conv2_block3_concat[0][0]        
                                                                 conv2_block4_2_conv[0][0]        
__________________________________________________________________________________________________
conv2_block5_0_bn (BatchNormali (None, 64, 64, 192)  768         conv2_block4_concat[0][0]        
__________________________________________________________________________________________________
conv2_block5_0_relu (Activation (None, 64, 64, 192)  0           conv2_block5_0_bn[0][0]          
__________________________________________________________________________________________________
conv2_block5_1_conv (Conv2D)    (None, 64, 64, 128)  24576       conv2_block5_0_relu[0][0]        
__________________________________________________________________________________________________
conv2_block5_1_bn (BatchNormali (None, 64, 64, 128)  512         conv2_block5_1_conv[0][0]        
__________________________________________________________________________________________________
conv2_block5_1_relu (Activation (None, 64, 64, 128)  0           conv2_block5_1_bn[0][0]          
__________________________________________________________________________________________________
conv2_block5_2_conv (Conv2D)    (None, 64, 64, 32)   36864       conv2_block5_1_relu[0][0]        
__________________________________________________________________________________________________
conv2_block5_concat (Concatenat (None, 64, 64, 224)  0           conv2_block4_concat[0][0]        
                                                                 conv2_block5_2_conv[0][0]        
__________________________________________________________________________________________________
conv2_block6_0_bn (BatchNormali (None, 64, 64, 224)  896         conv2_block5_concat[0][0]        
__________________________________________________________________________________________________
conv2_block6_0_relu (Activation (None, 64, 64, 224)  0           conv2_block6_0_bn[0][0]          
__________________________________________________________________________________________________
conv2_block6_1_conv (Conv2D)    (None, 64, 64, 128)  28672       conv2_block6_0_relu[0][0]        
__________________________________________________________________________________________________
conv2_block6_1_bn (BatchNormali (None, 64, 64, 128)  512         conv2_block6_1_conv[0][0]        
__________________________________________________________________________________________________
conv2_block6_1_relu (Activation (None, 64, 64, 128)  0           conv2_block6_1_bn[0][0]          
__________________________________________________________________________________________________
conv2_block6_2_conv (Conv2D)    (None, 64, 64, 32)   36864       conv2_block6_1_relu[0][0]        
__________________________________________________________________________________________________
conv2_block6_concat (Concatenat (None, 64, 64, 256)  0           conv2_block5_concat[0][0]        
                                                                 conv2_block6_2_conv[0][0]        
__________________________________________________________________________________________________
pool2_bn (BatchNormalization)   (None, 64, 64, 256)  1024        conv2_block6_concat[0][0]        
__________________________________________________________________________________________________
pool2_relu (Activation)         (None, 64, 64, 256)  0           pool2_bn[0][0]                   
__________________________________________________________________________________________________
pool2_conv (Conv2D)             (None, 64, 64, 128)  32768       pool2_relu[0][0]                 
__________________________________________________________________________________________________
pool2_pool (AveragePooling2D)   (None, 32, 32, 128)  0           pool2_conv[0][0]                 
__________________________________________________________________________________________________
conv3_block1_0_bn (BatchNormali (None, 32, 32, 128)  512         pool2_pool[0][0]                 
__________________________________________________________________________________________________
conv3_block1_0_relu (Activation (None, 32, 32, 128)  0           conv3_block1_0_bn[0][0]          
__________________________________________________________________________________________________
conv3_block1_1_conv (Conv2D)    (None, 32, 32, 128)  16384       conv3_block1_0_relu[0][0]        
__________________________________________________________________________________________________
conv3_block1_1_bn (BatchNormali (None, 32, 32, 128)  512         conv3_block1_1_conv[0][0]        
__________________________________________________________________________________________________
conv3_block1_1_relu (Activation (None, 32, 32, 128)  0           conv3_block1_1_bn[0][0]          
__________________________________________________________________________________________________
conv3_block1_2_conv (Conv2D)    (None, 32, 32, 32)   36864       conv3_block1_1_relu[0][0]        
__________________________________________________________________________________________________
conv3_block1_concat (Concatenat (None, 32, 32, 160)  0           pool2_pool[0][0]                 
                                                                 conv3_block1_2_conv[0][0]        
__________________________________________________________________________________________________
conv3_block2_0_bn (BatchNormali (None, 32, 32, 160)  640         conv3_block1_concat[0][0]        
__________________________________________________________________________________________________
conv3_block2_0_relu (Activation (None, 32, 32, 160)  0           conv3_block2_0_bn[0][0]          
__________________________________________________________________________________________________
conv3_block2_1_conv (Conv2D)    (None, 32, 32, 128)  20480       conv3_block2_0_relu[0][0]        
__________________________________________________________________________________________________
conv3_block2_1_bn (BatchNormali (None, 32, 32, 128)  512         conv3_block2_1_conv[0][0]        
__________________________________________________________________________________________________
conv3_block2_1_relu (Activation (None, 32, 32, 128)  0           conv3_block2_1_bn[0][0]          
__________________________________________________________________________________________________
conv3_block2_2_conv (Conv2D)    (None, 32, 32, 32)   36864       conv3_block2_1_relu[0][0]        
__________________________________________________________________________________________________
conv3_block2_concat (Concatenat (None, 32, 32, 192)  0           conv3_block1_concat[0][0]        
                                                                 conv3_block2_2_conv[0][0]        
__________________________________________________________________________________________________
conv3_block3_0_bn (BatchNormali (None, 32, 32, 192)  768         conv3_block2_concat[0][0]        
__________________________________________________________________________________________________
conv3_block3_0_relu (Activation (None, 32, 32, 192)  0           conv3_block3_0_bn[0][0]          
__________________________________________________________________________________________________
conv3_block3_1_conv (Conv2D)    (None, 32, 32, 128)  24576       conv3_block3_0_relu[0][0]        
__________________________________________________________________________________________________
conv3_block3_1_bn (BatchNormali (None, 32, 32, 128)  512         conv3_block3_1_conv[0][0]        
__________________________________________________________________________________________________
conv3_block3_1_relu (Activation (None, 32, 32, 128)  0           conv3_block3_1_bn[0][0]          
__________________________________________________________________________________________________
conv3_block3_2_conv (Conv2D)    (None, 32, 32, 32)   36864       conv3_block3_1_relu[0][0]        
__________________________________________________________________________________________________
conv3_block3_concat (Concatenat (None, 32, 32, 224)  0           conv3_block2_concat[0][0]        
                                                                 conv3_block3_2_conv[0][0]        
__________________________________________________________________________________________________
conv3_block4_0_bn (BatchNormali (None, 32, 32, 224)  896         conv3_block3_concat[0][0]        
__________________________________________________________________________________________________
conv3_block4_0_relu (Activation (None, 32, 32, 224)  0           conv3_block4_0_bn[0][0]          
__________________________________________________________________________________________________
conv3_block4_1_conv (Conv2D)    (None, 32, 32, 128)  28672       conv3_block4_0_relu[0][0]        
__________________________________________________________________________________________________
conv3_block4_1_bn (BatchNormali (None, 32, 32, 128)  512         conv3_block4_1_conv[0][0]        
__________________________________________________________________________________________________
conv3_block4_1_relu (Activation (None, 32, 32, 128)  0           conv3_block4_1_bn[0][0]          
__________________________________________________________________________________________________
conv3_block4_2_conv (Conv2D)    (None, 32, 32, 32)   36864       conv3_block4_1_relu[0][0]        
__________________________________________________________________________________________________
conv3_block4_concat (Concatenat (None, 32, 32, 256)  0           conv3_block3_concat[0][0]        
                                                                 conv3_block4_2_conv[0][0]        
__________________________________________________________________________________________________
conv3_block5_0_bn (BatchNormali (None, 32, 32, 256)  1024        conv3_block4_concat[0][0]        
__________________________________________________________________________________________________
conv3_block5_0_relu (Activation (None, 32, 32, 256)  0           conv3_block5_0_bn[0][0]          
__________________________________________________________________________________________________
conv3_block5_1_conv (Conv2D)    (None, 32, 32, 128)  32768       conv3_block5_0_relu[0][0]        
__________________________________________________________________________________________________
conv3_block5_1_bn (BatchNormali (None, 32, 32, 128)  512         conv3_block5_1_conv[0][0]        
__________________________________________________________________________________________________
conv3_block5_1_relu (Activation (None, 32, 32, 128)  0           conv3_block5_1_bn[0][0]          
__________________________________________________________________________________________________
conv3_block5_2_conv (Conv2D)    (None, 32, 32, 32)   36864       conv3_block5_1_relu[0][0]        
__________________________________________________________________________________________________
conv3_block5_concat (Concatenat (None, 32, 32, 288)  0           conv3_block4_concat[0][0]        
                                                                 conv3_block5_2_conv[0][0]        
__________________________________________________________________________________________________
conv3_block6_0_bn (BatchNormali (None, 32, 32, 288)  1152        conv3_block5_concat[0][0]        
__________________________________________________________________________________________________
conv3_block6_0_relu (Activation (None, 32, 32, 288)  0           conv3_block6_0_bn[0][0]          
__________________________________________________________________________________________________
conv3_block6_1_conv (Conv2D)    (None, 32, 32, 128)  36864       conv3_block6_0_relu[0][0]        
__________________________________________________________________________________________________
conv3_block6_1_bn (BatchNormali (None, 32, 32, 128)  512         conv3_block6_1_conv[0][0]        
__________________________________________________________________________________________________
conv3_block6_1_relu (Activation (None, 32, 32, 128)  0           conv3_block6_1_bn[0][0]          
__________________________________________________________________________________________________
conv3_block6_2_conv (Conv2D)    (None, 32, 32, 32)   36864       conv3_block6_1_relu[0][0]        
__________________________________________________________________________________________________
conv3_block6_concat (Concatenat (None, 32, 32, 320)  0           conv3_block5_concat[0][0]        
                                                                 conv3_block6_2_conv[0][0]        
__________________________________________________________________________________________________
conv3_block7_0_bn (BatchNormali (None, 32, 32, 320)  1280        conv3_block6_concat[0][0]        
__________________________________________________________________________________________________
conv3_block7_0_relu (Activation (None, 32, 32, 320)  0           conv3_block7_0_bn[0][0]          
__________________________________________________________________________________________________
conv3_block7_1_conv (Conv2D)    (None, 32, 32, 128)  40960       conv3_block7_0_relu[0][0]        
__________________________________________________________________________________________________
conv3_block7_1_bn (BatchNormali (None, 32, 32, 128)  512         conv3_block7_1_conv[0][0]        
__________________________________________________________________________________________________
conv3_block7_1_relu (Activation (None, 32, 32, 128)  0           conv3_block7_1_bn[0][0]          
__________________________________________________________________________________________________
conv3_block7_2_conv (Conv2D)    (None, 32, 32, 32)   36864       conv3_block7_1_relu[0][0]        
__________________________________________________________________________________________________
conv3_block7_concat (Concatenat (None, 32, 32, 352)  0           conv3_block6_concat[0][0]        
                                                                 conv3_block7_2_conv[0][0]        
__________________________________________________________________________________________________
conv3_block8_0_bn (BatchNormali (None, 32, 32, 352)  1408        conv3_block7_concat[0][0]        
__________________________________________________________________________________________________
conv3_block8_0_relu (Activation (None, 32, 32, 352)  0           conv3_block8_0_bn[0][0]          
__________________________________________________________________________________________________
conv3_block8_1_conv (Conv2D)    (None, 32, 32, 128)  45056       conv3_block8_0_relu[0][0]        
__________________________________________________________________________________________________
conv3_block8_1_bn (BatchNormali (None, 32, 32, 128)  512         conv3_block8_1_conv[0][0]        
__________________________________________________________________________________________________
conv3_block8_1_relu (Activation (None, 32, 32, 128)  0           conv3_block8_1_bn[0][0]          
__________________________________________________________________________________________________
conv3_block8_2_conv (Conv2D)    (None, 32, 32, 32)   36864       conv3_block8_1_relu[0][0]        
__________________________________________________________________________________________________
conv3_block8_concat (Concatenat (None, 32, 32, 384)  0           conv3_block7_concat[0][0]        
                                                                 conv3_block8_2_conv[0][0]        
__________________________________________________________________________________________________
conv3_block9_0_bn (BatchNormali (None, 32, 32, 384)  1536        conv3_block8_concat[0][0]        
__________________________________________________________________________________________________
conv3_block9_0_relu (Activation (None, 32, 32, 384)  0           conv3_block9_0_bn[0][0]          
__________________________________________________________________________________________________
conv3_block9_1_conv (Conv2D)    (None, 32, 32, 128)  49152       conv3_block9_0_relu[0][0]        
__________________________________________________________________________________________________
conv3_block9_1_bn (BatchNormali (None, 32, 32, 128)  512         conv3_block9_1_conv[0][0]        
__________________________________________________________________________________________________
conv3_block9_1_relu (Activation (None, 32, 32, 128)  0           conv3_block9_1_bn[0][0]          
__________________________________________________________________________________________________
conv3_block9_2_conv (Conv2D)    (None, 32, 32, 32)   36864       conv3_block9_1_relu[0][0]        
__________________________________________________________________________________________________
conv3_block9_concat (Concatenat (None, 32, 32, 416)  0           conv3_block8_concat[0][0]        
                                                                 conv3_block9_2_conv[0][0]        
__________________________________________________________________________________________________
conv3_block10_0_bn (BatchNormal (None, 32, 32, 416)  1664        conv3_block9_concat[0][0]        
__________________________________________________________________________________________________
conv3_block10_0_relu (Activatio (None, 32, 32, 416)  0           conv3_block10_0_bn[0][0]         
__________________________________________________________________________________________________
conv3_block10_1_conv (Conv2D)   (None, 32, 32, 128)  53248       conv3_block10_0_relu[0][0]       
__________________________________________________________________________________________________
conv3_block10_1_bn (BatchNormal (None, 32, 32, 128)  512         conv3_block10_1_conv[0][0]       
__________________________________________________________________________________________________
conv3_block10_1_relu (Activatio (None, 32, 32, 128)  0           conv3_block10_1_bn[0][0]         
__________________________________________________________________________________________________
conv3_block10_2_conv (Conv2D)   (None, 32, 32, 32)   36864       conv3_block10_1_relu[0][0]       
__________________________________________________________________________________________________
conv3_block10_concat (Concatena (None, 32, 32, 448)  0           conv3_block9_concat[0][0]        
                                                                 conv3_block10_2_conv[0][0]       
__________________________________________________________________________________________________
conv3_block11_0_bn (BatchNormal (None, 32, 32, 448)  1792        conv3_block10_concat[0][0]       
__________________________________________________________________________________________________
conv3_block11_0_relu (Activatio (None, 32, 32, 448)  0           conv3_block11_0_bn[0][0]         
__________________________________________________________________________________________________
conv3_block11_1_conv (Conv2D)   (None, 32, 32, 128)  57344       conv3_block11_0_relu[0][0]       
__________________________________________________________________________________________________
conv3_block11_1_bn (BatchNormal (None, 32, 32, 128)  512         conv3_block11_1_conv[0][0]       
__________________________________________________________________________________________________
conv3_block11_1_relu (Activatio (None, 32, 32, 128)  0           conv3_block11_1_bn[0][0]         
__________________________________________________________________________________________________
conv3_block11_2_conv (Conv2D)   (None, 32, 32, 32)   36864       conv3_block11_1_relu[0][0]       
__________________________________________________________________________________________________
conv3_block11_concat (Concatena (None, 32, 32, 480)  0           conv3_block10_concat[0][0]       
                                                                 conv3_block11_2_conv[0][0]       
__________________________________________________________________________________________________
conv3_block12_0_bn (BatchNormal (None, 32, 32, 480)  1920        conv3_block11_concat[0][0]       
__________________________________________________________________________________________________
conv3_block12_0_relu (Activatio (None, 32, 32, 480)  0           conv3_block12_0_bn[0][0]         
__________________________________________________________________________________________________
conv3_block12_1_conv (Conv2D)   (None, 32, 32, 128)  61440       conv3_block12_0_relu[0][0]       
__________________________________________________________________________________________________
conv3_block12_1_bn (BatchNormal (None, 32, 32, 128)  512         conv3_block12_1_conv[0][0]       
__________________________________________________________________________________________________
conv3_block12_1_relu (Activatio (None, 32, 32, 128)  0           conv3_block12_1_bn[0][0]         
__________________________________________________________________________________________________
conv3_block12_2_conv (Conv2D)   (None, 32, 32, 32)   36864       conv3_block12_1_relu[0][0]       
__________________________________________________________________________________________________
conv3_block12_concat (Concatena (None, 32, 32, 512)  0           conv3_block11_concat[0][0]       
                                                                 conv3_block12_2_conv[0][0]       
__________________________________________________________________________________________________
pool3_bn (BatchNormalization)   (None, 32, 32, 512)  2048        conv3_block12_concat[0][0]       
__________________________________________________________________________________________________
pool3_relu (Activation)         (None, 32, 32, 512)  0           pool3_bn[0][0]                   
__________________________________________________________________________________________________
pool3_conv (Conv2D)             (None, 32, 32, 256)  131072      pool3_relu[0][0]                 
__________________________________________________________________________________________________
pool3_pool (AveragePooling2D)   (None, 16, 16, 256)  0           pool3_conv[0][0]                 
__________________________________________________________________________________________________
conv4_block1_0_bn (BatchNormali (None, 16, 16, 256)  1024        pool3_pool[0][0]                 
__________________________________________________________________________________________________
conv4_block1_0_relu (Activation (None, 16, 16, 256)  0           conv4_block1_0_bn[0][0]          
__________________________________________________________________________________________________
conv4_block1_1_conv (Conv2D)    (None, 16, 16, 128)  32768       conv4_block1_0_relu[0][0]        
__________________________________________________________________________________________________
conv4_block1_1_bn (BatchNormali (None, 16, 16, 128)  512         conv4_block1_1_conv[0][0]        
__________________________________________________________________________________________________
conv4_block1_1_relu (Activation (None, 16, 16, 128)  0           conv4_block1_1_bn[0][0]          
__________________________________________________________________________________________________
conv4_block1_2_conv (Conv2D)    (None, 16, 16, 32)   36864       conv4_block1_1_relu[0][0]        
__________________________________________________________________________________________________
conv4_block1_concat (Concatenat (None, 16, 16, 288)  0           pool3_pool[0][0]                 
                                                                 conv4_block1_2_conv[0][0]        
__________________________________________________________________________________________________
conv4_block2_0_bn (BatchNormali (None, 16, 16, 288)  1152        conv4_block1_concat[0][0]        
__________________________________________________________________________________________________
conv4_block2_0_relu (Activation (None, 16, 16, 288)  0           conv4_block2_0_bn[0][0]          
__________________________________________________________________________________________________
conv4_block2_1_conv (Conv2D)    (None, 16, 16, 128)  36864       conv4_block2_0_relu[0][0]        
__________________________________________________________________________________________________
conv4_block2_1_bn (BatchNormali (None, 16, 16, 128)  512         conv4_block2_1_conv[0][0]        
__________________________________________________________________________________________________
conv4_block2_1_relu (Activation (None, 16, 16, 128)  0           conv4_block2_1_bn[0][0]          
__________________________________________________________________________________________________
conv4_block2_2_conv (Conv2D)    (None, 16, 16, 32)   36864       conv4_block2_1_relu[0][0]        
__________________________________________________________________________________________________
conv4_block2_concat (Concatenat (None, 16, 16, 320)  0           conv4_block1_concat[0][0]        
                                                                 conv4_block2_2_conv[0][0]        
__________________________________________________________________________________________________
conv4_block3_0_bn (BatchNormali (None, 16, 16, 320)  1280        conv4_block2_concat[0][0]        
__________________________________________________________________________________________________
conv4_block3_0_relu (Activation (None, 16, 16, 320)  0           conv4_block3_0_bn[0][0]          
__________________________________________________________________________________________________
conv4_block3_1_conv (Conv2D)    (None, 16, 16, 128)  40960       conv4_block3_0_relu[0][0]        
__________________________________________________________________________________________________
conv4_block3_1_bn (BatchNormali (None, 16, 16, 128)  512         conv4_block3_1_conv[0][0]        
__________________________________________________________________________________________________
conv4_block3_1_relu (Activation (None, 16, 16, 128)  0           conv4_block3_1_bn[0][0]          
__________________________________________________________________________________________________
conv4_block3_2_conv (Conv2D)    (None, 16, 16, 32)   36864       conv4_block3_1_relu[0][0]        
__________________________________________________________________________________________________
conv4_block3_concat (Concatenat (None, 16, 16, 352)  0           conv4_block2_concat[0][0]        
                                                                 conv4_block3_2_conv[0][0]        
__________________________________________________________________________________________________
conv4_block4_0_bn (BatchNormali (None, 16, 16, 352)  1408        conv4_block3_concat[0][0]        
__________________________________________________________________________________________________
conv4_block4_0_relu (Activation (None, 16, 16, 352)  0           conv4_block4_0_bn[0][0]          
__________________________________________________________________________________________________
conv4_block4_1_conv (Conv2D)    (None, 16, 16, 128)  45056       conv4_block4_0_relu[0][0]        
__________________________________________________________________________________________________
conv4_block4_1_bn (BatchNormali (None, 16, 16, 128)  512         conv4_block4_1_conv[0][0]        
__________________________________________________________________________________________________
conv4_block4_1_relu (Activation (None, 16, 16, 128)  0           conv4_block4_1_bn[0][0]          
__________________________________________________________________________________________________
conv4_block4_2_conv (Conv2D)    (None, 16, 16, 32)   36864       conv4_block4_1_relu[0][0]        
__________________________________________________________________________________________________
conv4_block4_concat (Concatenat (None, 16, 16, 384)  0           conv4_block3_concat[0][0]        
                                                                 conv4_block4_2_conv[0][0]        
__________________________________________________________________________________________________
conv4_block5_0_bn (BatchNormali (None, 16, 16, 384)  1536        conv4_block4_concat[0][0]        
__________________________________________________________________________________________________
conv4_block5_0_relu (Activation (None, 16, 16, 384)  0           conv4_block5_0_bn[0][0]          
__________________________________________________________________________________________________
conv4_block5_1_conv (Conv2D)    (None, 16, 16, 128)  49152       conv4_block5_0_relu[0][0]        
__________________________________________________________________________________________________
conv4_block5_1_bn (BatchNormali (None, 16, 16, 128)  512         conv4_block5_1_conv[0][0]        
__________________________________________________________________________________________________
conv4_block5_1_relu (Activation (None, 16, 16, 128)  0           conv4_block5_1_bn[0][0]          
__________________________________________________________________________________________________
conv4_block5_2_conv (Conv2D)    (None, 16, 16, 32)   36864       conv4_block5_1_relu[0][0]        
__________________________________________________________________________________________________
conv4_block5_concat (Concatenat (None, 16, 16, 416)  0           conv4_block4_concat[0][0]        
                                                                 conv4_block5_2_conv[0][0]        
__________________________________________________________________________________________________
conv4_block6_0_bn (BatchNormali (None, 16, 16, 416)  1664        conv4_block5_concat[0][0]        
__________________________________________________________________________________________________
conv4_block6_0_relu (Activation (None, 16, 16, 416)  0           conv4_block6_0_bn[0][0]          
__________________________________________________________________________________________________
conv4_block6_1_conv (Conv2D)    (None, 16, 16, 128)  53248       conv4_block6_0_relu[0][0]        
__________________________________________________________________________________________________
conv4_block6_1_bn (BatchNormali (None, 16, 16, 128)  512         conv4_block6_1_conv[0][0]        
__________________________________________________________________________________________________
conv4_block6_1_relu (Activation (None, 16, 16, 128)  0           conv4_block6_1_bn[0][0]          
__________________________________________________________________________________________________
conv4_block6_2_conv (Conv2D)    (None, 16, 16, 32)   36864       conv4_block6_1_relu[0][0]        
__________________________________________________________________________________________________
conv4_block6_concat (Concatenat (None, 16, 16, 448)  0           conv4_block5_concat[0][0]        
                                                                 conv4_block6_2_conv[0][0]        
__________________________________________________________________________________________________
conv4_block7_0_bn (BatchNormali (None, 16, 16, 448)  1792        conv4_block6_concat[0][0]        
__________________________________________________________________________________________________
conv4_block7_0_relu (Activation (None, 16, 16, 448)  0           conv4_block7_0_bn[0][0]          
__________________________________________________________________________________________________
conv4_block7_1_conv (Conv2D)    (None, 16, 16, 128)  57344       conv4_block7_0_relu[0][0]        
__________________________________________________________________________________________________
conv4_block7_1_bn (BatchNormali (None, 16, 16, 128)  512         conv4_block7_1_conv[0][0]        
__________________________________________________________________________________________________
conv4_block7_1_relu (Activation (None, 16, 16, 128)  0           conv4_block7_1_bn[0][0]          
__________________________________________________________________________________________________
conv4_block7_2_conv (Conv2D)    (None, 16, 16, 32)   36864       conv4_block7_1_relu[0][0]        
__________________________________________________________________________________________________
conv4_block7_concat (Concatenat (None, 16, 16, 480)  0           conv4_block6_concat[0][0]        
                                                                 conv4_block7_2_conv[0][0]        
__________________________________________________________________________________________________
conv4_block8_0_bn (BatchNormali (None, 16, 16, 480)  1920        conv4_block7_concat[0][0]        
__________________________________________________________________________________________________
conv4_block8_0_relu (Activation (None, 16, 16, 480)  0           conv4_block8_0_bn[0][0]          
__________________________________________________________________________________________________
conv4_block8_1_conv (Conv2D)    (None, 16, 16, 128)  61440       conv4_block8_0_relu[0][0]        
__________________________________________________________________________________________________
conv4_block8_1_bn (BatchNormali (None, 16, 16, 128)  512         conv4_block8_1_conv[0][0]        
__________________________________________________________________________________________________
conv4_block8_1_relu (Activation (None, 16, 16, 128)  0           conv4_block8_1_bn[0][0]          
__________________________________________________________________________________________________
conv4_block8_2_conv (Conv2D)    (None, 16, 16, 32)   36864       conv4_block8_1_relu[0][0]        
__________________________________________________________________________________________________
conv4_block8_concat (Concatenat (None, 16, 16, 512)  0           conv4_block7_concat[0][0]        
                                                                 conv4_block8_2_conv[0][0]        
__________________________________________________________________________________________________
conv4_block9_0_bn (BatchNormali (None, 16, 16, 512)  2048        conv4_block8_concat[0][0]        
__________________________________________________________________________________________________
conv4_block9_0_relu (Activation (None, 16, 16, 512)  0           conv4_block9_0_bn[0][0]          
__________________________________________________________________________________________________
conv4_block9_1_conv (Conv2D)    (None, 16, 16, 128)  65536       conv4_block9_0_relu[0][0]        
__________________________________________________________________________________________________
conv4_block9_1_bn (BatchNormali (None, 16, 16, 128)  512         conv4_block9_1_conv[0][0]        
__________________________________________________________________________________________________
conv4_block9_1_relu (Activation (None, 16, 16, 128)  0           conv4_block9_1_bn[0][0]          
__________________________________________________________________________________________________
conv4_block9_2_conv (Conv2D)    (None, 16, 16, 32)   36864       conv4_block9_1_relu[0][0]        
__________________________________________________________________________________________________
conv4_block9_concat (Concatenat (None, 16, 16, 544)  0           conv4_block8_concat[0][0]        
                                                                 conv4_block9_2_conv[0][0]        
__________________________________________________________________________________________________
conv4_block10_0_bn (BatchNormal (None, 16, 16, 544)  2176        conv4_block9_concat[0][0]        
__________________________________________________________________________________________________
conv4_block10_0_relu (Activatio (None, 16, 16, 544)  0           conv4_block10_0_bn[0][0]         
__________________________________________________________________________________________________
conv4_block10_1_conv (Conv2D)   (None, 16, 16, 128)  69632       conv4_block10_0_relu[0][0]       
__________________________________________________________________________________________________
conv4_block10_1_bn (BatchNormal (None, 16, 16, 128)  512         conv4_block10_1_conv[0][0]       
__________________________________________________________________________________________________
conv4_block10_1_relu (Activatio (None, 16, 16, 128)  0           conv4_block10_1_bn[0][0]         
__________________________________________________________________________________________________
conv4_block10_2_conv (Conv2D)   (None, 16, 16, 32)   36864       conv4_block10_1_relu[0][0]       
__________________________________________________________________________________________________
conv4_block10_concat (Concatena (None, 16, 16, 576)  0           conv4_block9_concat[0][0]        
                                                                 conv4_block10_2_conv[0][0]       
__________________________________________________________________________________________________
conv4_block11_0_bn (BatchNormal (None, 16, 16, 576)  2304        conv4_block10_concat[0][0]       
__________________________________________________________________________________________________
conv4_block11_0_relu (Activatio (None, 16, 16, 576)  0           conv4_block11_0_bn[0][0]         
__________________________________________________________________________________________________
conv4_block11_1_conv (Conv2D)   (None, 16, 16, 128)  73728       conv4_block11_0_relu[0][0]       
__________________________________________________________________________________________________
conv4_block11_1_bn (BatchNormal (None, 16, 16, 128)  512         conv4_block11_1_conv[0][0]       
__________________________________________________________________________________________________
conv4_block11_1_relu (Activatio (None, 16, 16, 128)  0           conv4_block11_1_bn[0][0]         
__________________________________________________________________________________________________
conv4_block11_2_conv (Conv2D)   (None, 16, 16, 32)   36864       conv4_block11_1_relu[0][0]       
__________________________________________________________________________________________________
conv4_block11_concat (Concatena (None, 16, 16, 608)  0           conv4_block10_concat[0][0]       
                                                                 conv4_block11_2_conv[0][0]       
__________________________________________________________________________________________________
conv4_block12_0_bn (BatchNormal (None, 16, 16, 608)  2432        conv4_block11_concat[0][0]       
__________________________________________________________________________________________________
conv4_block12_0_relu (Activatio (None, 16, 16, 608)  0           conv4_block12_0_bn[0][0]         
__________________________________________________________________________________________________
conv4_block12_1_conv (Conv2D)   (None, 16, 16, 128)  77824       conv4_block12_0_relu[0][0]       
__________________________________________________________________________________________________
conv4_block12_1_bn (BatchNormal (None, 16, 16, 128)  512         conv4_block12_1_conv[0][0]       
__________________________________________________________________________________________________
conv4_block12_1_relu (Activatio (None, 16, 16, 128)  0           conv4_block12_1_bn[0][0]         
__________________________________________________________________________________________________
conv4_block12_2_conv (Conv2D)   (None, 16, 16, 32)   36864       conv4_block12_1_relu[0][0]       
__________________________________________________________________________________________________
conv4_block12_concat (Concatena (None, 16, 16, 640)  0           conv4_block11_concat[0][0]       
                                                                 conv4_block12_2_conv[0][0]       
__________________________________________________________________________________________________
conv4_block13_0_bn (BatchNormal (None, 16, 16, 640)  2560        conv4_block12_concat[0][0]       
__________________________________________________________________________________________________
conv4_block13_0_relu (Activatio (None, 16, 16, 640)  0           conv4_block13_0_bn[0][0]         
__________________________________________________________________________________________________
conv4_block13_1_conv (Conv2D)   (None, 16, 16, 128)  81920       conv4_block13_0_relu[0][0]       
__________________________________________________________________________________________________
conv4_block13_1_bn (BatchNormal (None, 16, 16, 128)  512         conv4_block13_1_conv[0][0]       
__________________________________________________________________________________________________
conv4_block13_1_relu (Activatio (None, 16, 16, 128)  0           conv4_block13_1_bn[0][0]         
__________________________________________________________________________________________________
conv4_block13_2_conv (Conv2D)   (None, 16, 16, 32)   36864       conv4_block13_1_relu[0][0]       
__________________________________________________________________________________________________
conv4_block13_concat (Concatena (None, 16, 16, 672)  0           conv4_block12_concat[0][0]       
                                                                 conv4_block13_2_conv[0][0]       
__________________________________________________________________________________________________
conv4_block14_0_bn (BatchNormal (None, 16, 16, 672)  2688        conv4_block13_concat[0][0]       
__________________________________________________________________________________________________
conv4_block14_0_relu (Activatio (None, 16, 16, 672)  0           conv4_block14_0_bn[0][0]         
__________________________________________________________________________________________________
conv4_block14_1_conv (Conv2D)   (None, 16, 16, 128)  86016       conv4_block14_0_relu[0][0]       
__________________________________________________________________________________________________
conv4_block14_1_bn (BatchNormal (None, 16, 16, 128)  512         conv4_block14_1_conv[0][0]       
__________________________________________________________________________________________________
conv4_block14_1_relu (Activatio (None, 16, 16, 128)  0           conv4_block14_1_bn[0][0]         
__________________________________________________________________________________________________
conv4_block14_2_conv (Conv2D)   (None, 16, 16, 32)   36864       conv4_block14_1_relu[0][0]       
__________________________________________________________________________________________________
conv4_block14_concat (Concatena (None, 16, 16, 704)  0           conv4_block13_concat[0][0]       
                                                                 conv4_block14_2_conv[0][0]       
__________________________________________________________________________________________________
conv4_block15_0_bn (BatchNormal (None, 16, 16, 704)  2816        conv4_block14_concat[0][0]       
__________________________________________________________________________________________________
conv4_block15_0_relu (Activatio (None, 16, 16, 704)  0           conv4_block15_0_bn[0][0]         
__________________________________________________________________________________________________
conv4_block15_1_conv (Conv2D)   (None, 16, 16, 128)  90112       conv4_block15_0_relu[0][0]       
__________________________________________________________________________________________________
conv4_block15_1_bn (BatchNormal (None, 16, 16, 128)  512         conv4_block15_1_conv[0][0]       
__________________________________________________________________________________________________
conv4_block15_1_relu (Activatio (None, 16, 16, 128)  0           conv4_block15_1_bn[0][0]         
__________________________________________________________________________________________________
conv4_block15_2_conv (Conv2D)   (None, 16, 16, 32)   36864       conv4_block15_1_relu[0][0]       
__________________________________________________________________________________________________
conv4_block15_concat (Concatena (None, 16, 16, 736)  0           conv4_block14_concat[0][0]       
                                                                 conv4_block15_2_conv[0][0]       
__________________________________________________________________________________________________
conv4_block16_0_bn (BatchNormal (None, 16, 16, 736)  2944        conv4_block15_concat[0][0]       
__________________________________________________________________________________________________
conv4_block16_0_relu (Activatio (None, 16, 16, 736)  0           conv4_block16_0_bn[0][0]         
__________________________________________________________________________________________________
conv4_block16_1_conv (Conv2D)   (None, 16, 16, 128)  94208       conv4_block16_0_relu[0][0]       
__________________________________________________________________________________________________
conv4_block16_1_bn (BatchNormal (None, 16, 16, 128)  512         conv4_block16_1_conv[0][0]       
__________________________________________________________________________________________________
conv4_block16_1_relu (Activatio (None, 16, 16, 128)  0           conv4_block16_1_bn[0][0]         
__________________________________________________________________________________________________
conv4_block16_2_conv (Conv2D)   (None, 16, 16, 32)   36864       conv4_block16_1_relu[0][0]       
__________________________________________________________________________________________________
conv4_block16_concat (Concatena (None, 16, 16, 768)  0           conv4_block15_concat[0][0]       
                                                                 conv4_block16_2_conv[0][0]       
__________________________________________________________________________________________________
conv4_block17_0_bn (BatchNormal (None, 16, 16, 768)  3072        conv4_block16_concat[0][0]       
__________________________________________________________________________________________________
conv4_block17_0_relu (Activatio (None, 16, 16, 768)  0           conv4_block17_0_bn[0][0]         
__________________________________________________________________________________________________
conv4_block17_1_conv (Conv2D)   (None, 16, 16, 128)  98304       conv4_block17_0_relu[0][0]       
__________________________________________________________________________________________________
conv4_block17_1_bn (BatchNormal (None, 16, 16, 128)  512         conv4_block17_1_conv[0][0]       
__________________________________________________________________________________________________
conv4_block17_1_relu (Activatio (None, 16, 16, 128)  0           conv4_block17_1_bn[0][0]         
__________________________________________________________________________________________________
conv4_block17_2_conv (Conv2D)   (None, 16, 16, 32)   36864       conv4_block17_1_relu[0][0]       
__________________________________________________________________________________________________
conv4_block17_concat (Concatena (None, 16, 16, 800)  0           conv4_block16_concat[0][0]       
                                                                 conv4_block17_2_conv[0][0]       
__________________________________________________________________________________________________
conv4_block18_0_bn (BatchNormal (None, 16, 16, 800)  3200        conv4_block17_concat[0][0]       
__________________________________________________________________________________________________
conv4_block18_0_relu (Activatio (None, 16, 16, 800)  0           conv4_block18_0_bn[0][0]         
__________________________________________________________________________________________________
conv4_block18_1_conv (Conv2D)   (None, 16, 16, 128)  102400      conv4_block18_0_relu[0][0]       
__________________________________________________________________________________________________
conv4_block18_1_bn (BatchNormal (None, 16, 16, 128)  512         conv4_block18_1_conv[0][0]       
__________________________________________________________________________________________________
conv4_block18_1_relu (Activatio (None, 16, 16, 128)  0           conv4_block18_1_bn[0][0]         
__________________________________________________________________________________________________
conv4_block18_2_conv (Conv2D)   (None, 16, 16, 32)   36864       conv4_block18_1_relu[0][0]       
__________________________________________________________________________________________________
conv4_block18_concat (Concatena (None, 16, 16, 832)  0           conv4_block17_concat[0][0]       
                                                                 conv4_block18_2_conv[0][0]       
__________________________________________________________________________________________________
conv4_block19_0_bn (BatchNormal (None, 16, 16, 832)  3328        conv4_block18_concat[0][0]       
__________________________________________________________________________________________________
conv4_block19_0_relu (Activatio (None, 16, 16, 832)  0           conv4_block19_0_bn[0][0]         
__________________________________________________________________________________________________
conv4_block19_1_conv (Conv2D)   (None, 16, 16, 128)  106496      conv4_block19_0_relu[0][0]       
__________________________________________________________________________________________________
conv4_block19_1_bn (BatchNormal (None, 16, 16, 128)  512         conv4_block19_1_conv[0][0]       
__________________________________________________________________________________________________
conv4_block19_1_relu (Activatio (None, 16, 16, 128)  0           conv4_block19_1_bn[0][0]         
__________________________________________________________________________________________________
conv4_block19_2_conv (Conv2D)   (None, 16, 16, 32)   36864       conv4_block19_1_relu[0][0]       
__________________________________________________________________________________________________
conv4_block19_concat (Concatena (None, 16, 16, 864)  0           conv4_block18_concat[0][0]       
                                                                 conv4_block19_2_conv[0][0]       
__________________________________________________________________________________________________
conv4_block20_0_bn (BatchNormal (None, 16, 16, 864)  3456        conv4_block19_concat[0][0]       
__________________________________________________________________________________________________
conv4_block20_0_relu (Activatio (None, 16, 16, 864)  0           conv4_block20_0_bn[0][0]         
__________________________________________________________________________________________________
conv4_block20_1_conv (Conv2D)   (None, 16, 16, 128)  110592      conv4_block20_0_relu[0][0]       
__________________________________________________________________________________________________
conv4_block20_1_bn (BatchNormal (None, 16, 16, 128)  512         conv4_block20_1_conv[0][0]       
__________________________________________________________________________________________________
conv4_block20_1_relu (Activatio (None, 16, 16, 128)  0           conv4_block20_1_bn[0][0]         
__________________________________________________________________________________________________
conv4_block20_2_conv (Conv2D)   (None, 16, 16, 32)   36864       conv4_block20_1_relu[0][0]       
__________________________________________________________________________________________________
conv4_block20_concat (Concatena (None, 16, 16, 896)  0           conv4_block19_concat[0][0]       
                                                                 conv4_block20_2_conv[0][0]       
__________________________________________________________________________________________________
conv4_block21_0_bn (BatchNormal (None, 16, 16, 896)  3584        conv4_block20_concat[0][0]       
__________________________________________________________________________________________________
conv4_block21_0_relu (Activatio (None, 16, 16, 896)  0           conv4_block21_0_bn[0][0]         
__________________________________________________________________________________________________
conv4_block21_1_conv (Conv2D)   (None, 16, 16, 128)  114688      conv4_block21_0_relu[0][0]       
__________________________________________________________________________________________________
conv4_block21_1_bn (BatchNormal (None, 16, 16, 128)  512         conv4_block21_1_conv[0][0]       
__________________________________________________________________________________________________
conv4_block21_1_relu (Activatio (None, 16, 16, 128)  0           conv4_block21_1_bn[0][0]         
__________________________________________________________________________________________________
conv4_block21_2_conv (Conv2D)   (None, 16, 16, 32)   36864       conv4_block21_1_relu[0][0]       
__________________________________________________________________________________________________
conv4_block21_concat (Concatena (None, 16, 16, 928)  0           conv4_block20_concat[0][0]       
                                                                 conv4_block21_2_conv[0][0]       
__________________________________________________________________________________________________
conv4_block22_0_bn (BatchNormal (None, 16, 16, 928)  3712        conv4_block21_concat[0][0]       
__________________________________________________________________________________________________
conv4_block22_0_relu (Activatio (None, 16, 16, 928)  0           conv4_block22_0_bn[0][0]         
__________________________________________________________________________________________________
conv4_block22_1_conv (Conv2D)   (None, 16, 16, 128)  118784      conv4_block22_0_relu[0][0]       
__________________________________________________________________________________________________
conv4_block22_1_bn (BatchNormal (None, 16, 16, 128)  512         conv4_block22_1_conv[0][0]       
__________________________________________________________________________________________________
conv4_block22_1_relu (Activatio (None, 16, 16, 128)  0           conv4_block22_1_bn[0][0]         
__________________________________________________________________________________________________
conv4_block22_2_conv (Conv2D)   (None, 16, 16, 32)   36864       conv4_block22_1_relu[0][0]       
__________________________________________________________________________________________________
conv4_block22_concat (Concatena (None, 16, 16, 960)  0           conv4_block21_concat[0][0]       
                                                                 conv4_block22_2_conv[0][0]       
__________________________________________________________________________________________________
conv4_block23_0_bn (BatchNormal (None, 16, 16, 960)  3840        conv4_block22_concat[0][0]       
__________________________________________________________________________________________________
conv4_block23_0_relu (Activatio (None, 16, 16, 960)  0           conv4_block23_0_bn[0][0]         
__________________________________________________________________________________________________
conv4_block23_1_conv (Conv2D)   (None, 16, 16, 128)  122880      conv4_block23_0_relu[0][0]       
__________________________________________________________________________________________________
conv4_block23_1_bn (BatchNormal (None, 16, 16, 128)  512         conv4_block23_1_conv[0][0]       
__________________________________________________________________________________________________
conv4_block23_1_relu (Activatio (None, 16, 16, 128)  0           conv4_block23_1_bn[0][0]         
__________________________________________________________________________________________________
conv4_block23_2_conv (Conv2D)   (None, 16, 16, 32)   36864       conv4_block23_1_relu[0][0]       
__________________________________________________________________________________________________
conv4_block23_concat (Concatena (None, 16, 16, 992)  0           conv4_block22_concat[0][0]       
                                                                 conv4_block23_2_conv[0][0]       
__________________________________________________________________________________________________
conv4_block24_0_bn (BatchNormal (None, 16, 16, 992)  3968        conv4_block23_concat[0][0]       
__________________________________________________________________________________________________
conv4_block24_0_relu (Activatio (None, 16, 16, 992)  0           conv4_block24_0_bn[0][0]         
__________________________________________________________________________________________________
conv4_block24_1_conv (Conv2D)   (None, 16, 16, 128)  126976      conv4_block24_0_relu[0][0]       
__________________________________________________________________________________________________
conv4_block24_1_bn (BatchNormal (None, 16, 16, 128)  512         conv4_block24_1_conv[0][0]       
__________________________________________________________________________________________________
conv4_block24_1_relu (Activatio (None, 16, 16, 128)  0           conv4_block24_1_bn[0][0]         
__________________________________________________________________________________________________
conv4_block24_2_conv (Conv2D)   (None, 16, 16, 32)   36864       conv4_block24_1_relu[0][0]       
__________________________________________________________________________________________________
conv4_block24_concat (Concatena (None, 16, 16, 1024) 0           conv4_block23_concat[0][0]       
                                                                 conv4_block24_2_conv[0][0]       
__________________________________________________________________________________________________
pool4_bn (BatchNormalization)   (None, 16, 16, 1024) 4096        conv4_block24_concat[0][0]       
__________________________________________________________________________________________________
pool4_relu (Activation)         (None, 16, 16, 1024) 0           pool4_bn[0][0]                   
__________________________________________________________________________________________________
pool4_conv (Conv2D)             (None, 16, 16, 512)  524288      pool4_relu[0][0]                 
__________________________________________________________________________________________________
pool4_pool (AveragePooling2D)   (None, 8, 8, 512)    0           pool4_conv[0][0]                 
__________________________________________________________________________________________________
conv5_block1_0_bn (BatchNormali (None, 8, 8, 512)    2048        pool4_pool[0][0]                 
__________________________________________________________________________________________________
conv5_block1_0_relu (Activation (None, 8, 8, 512)    0           conv5_block1_0_bn[0][0]          
__________________________________________________________________________________________________
conv5_block1_1_conv (Conv2D)    (None, 8, 8, 128)    65536       conv5_block1_0_relu[0][0]        
__________________________________________________________________________________________________
conv5_block1_1_bn (BatchNormali (None, 8, 8, 128)    512         conv5_block1_1_conv[0][0]        
__________________________________________________________________________________________________
conv5_block1_1_relu (Activation (None, 8, 8, 128)    0           conv5_block1_1_bn[0][0]          
__________________________________________________________________________________________________
conv5_block1_2_conv (Conv2D)    (None, 8, 8, 32)     36864       conv5_block1_1_relu[0][0]        
__________________________________________________________________________________________________
conv5_block1_concat (Concatenat (None, 8, 8, 544)    0           pool4_pool[0][0]                 
                                                                 conv5_block1_2_conv[0][0]        
__________________________________________________________________________________________________
conv5_block2_0_bn (BatchNormali (None, 8, 8, 544)    2176        conv5_block1_concat[0][0]        
__________________________________________________________________________________________________
conv5_block2_0_relu (Activation (None, 8, 8, 544)    0           conv5_block2_0_bn[0][0]          
__________________________________________________________________________________________________
conv5_block2_1_conv (Conv2D)    (None, 8, 8, 128)    69632       conv5_block2_0_relu[0][0]        
__________________________________________________________________________________________________
conv5_block2_1_bn (BatchNormali (None, 8, 8, 128)    512         conv5_block2_1_conv[0][0]        
__________________________________________________________________________________________________
conv5_block2_1_relu (Activation (None, 8, 8, 128)    0           conv5_block2_1_bn[0][0]          
__________________________________________________________________________________________________
conv5_block2_2_conv (Conv2D)    (None, 8, 8, 32)     36864       conv5_block2_1_relu[0][0]        
__________________________________________________________________________________________________
conv5_block2_concat (Concatenat (None, 8, 8, 576)    0           conv5_block1_concat[0][0]        
                                                                 conv5_block2_2_conv[0][0]        
__________________________________________________________________________________________________
conv5_block3_0_bn (BatchNormali (None, 8, 8, 576)    2304        conv5_block2_concat[0][0]        
__________________________________________________________________________________________________
conv5_block3_0_relu (Activation (None, 8, 8, 576)    0           conv5_block3_0_bn[0][0]          
__________________________________________________________________________________________________
conv5_block3_1_conv (Conv2D)    (None, 8, 8, 128)    73728       conv5_block3_0_relu[0][0]        
__________________________________________________________________________________________________
conv5_block3_1_bn (BatchNormali (None, 8, 8, 128)    512         conv5_block3_1_conv[0][0]        
__________________________________________________________________________________________________
conv5_block3_1_relu (Activation (None, 8, 8, 128)    0           conv5_block3_1_bn[0][0]          
__________________________________________________________________________________________________
conv5_block3_2_conv (Conv2D)    (None, 8, 8, 32)     36864       conv5_block3_1_relu[0][0]        
__________________________________________________________________________________________________
conv5_block3_concat (Concatenat (None, 8, 8, 608)    0           conv5_block2_concat[0][0]        
                                                                 conv5_block3_2_conv[0][0]        
__________________________________________________________________________________________________
conv5_block4_0_bn (BatchNormali (None, 8, 8, 608)    2432        conv5_block3_concat[0][0]        
__________________________________________________________________________________________________
conv5_block4_0_relu (Activation (None, 8, 8, 608)    0           conv5_block4_0_bn[0][0]          
__________________________________________________________________________________________________
conv5_block4_1_conv (Conv2D)    (None, 8, 8, 128)    77824       conv5_block4_0_relu[0][0]        
__________________________________________________________________________________________________
conv5_block4_1_bn (BatchNormali (None, 8, 8, 128)    512         conv5_block4_1_conv[0][0]        
__________________________________________________________________________________________________
conv5_block4_1_relu (Activation (None, 8, 8, 128)    0           conv5_block4_1_bn[0][0]          
__________________________________________________________________________________________________
conv5_block4_2_conv (Conv2D)    (None, 8, 8, 32)     36864       conv5_block4_1_relu[0][0]        
__________________________________________________________________________________________________
conv5_block4_concat (Concatenat (None, 8, 8, 640)    0           conv5_block3_concat[0][0]        
                                                                 conv5_block4_2_conv[0][0]        
__________________________________________________________________________________________________
conv5_block5_0_bn (BatchNormali (None, 8, 8, 640)    2560        conv5_block4_concat[0][0]        
__________________________________________________________________________________________________
conv5_block5_0_relu (Activation (None, 8, 8, 640)    0           conv5_block5_0_bn[0][0]          
__________________________________________________________________________________________________
conv5_block5_1_conv (Conv2D)    (None, 8, 8, 128)    81920       conv5_block5_0_relu[0][0]        
__________________________________________________________________________________________________
conv5_block5_1_bn (BatchNormali (None, 8, 8, 128)    512         conv5_block5_1_conv[0][0]        
__________________________________________________________________________________________________
conv5_block5_1_relu (Activation (None, 8, 8, 128)    0           conv5_block5_1_bn[0][0]          
__________________________________________________________________________________________________
conv5_block5_2_conv (Conv2D)    (None, 8, 8, 32)     36864       conv5_block5_1_relu[0][0]        
__________________________________________________________________________________________________
conv5_block5_concat (Concatenat (None, 8, 8, 672)    0           conv5_block4_concat[0][0]        
                                                                 conv5_block5_2_conv[0][0]        
__________________________________________________________________________________________________
conv5_block6_0_bn (BatchNormali (None, 8, 8, 672)    2688        conv5_block5_concat[0][0]        
__________________________________________________________________________________________________
conv5_block6_0_relu (Activation (None, 8, 8, 672)    0           conv5_block6_0_bn[0][0]          
__________________________________________________________________________________________________
conv5_block6_1_conv (Conv2D)    (None, 8, 8, 128)    86016       conv5_block6_0_relu[0][0]        
__________________________________________________________________________________________________
conv5_block6_1_bn (BatchNormali (None, 8, 8, 128)    512         conv5_block6_1_conv[0][0]        
__________________________________________________________________________________________________
conv5_block6_1_relu (Activation (None, 8, 8, 128)    0           conv5_block6_1_bn[0][0]          
__________________________________________________________________________________________________
conv5_block6_2_conv (Conv2D)    (None, 8, 8, 32)     36864       conv5_block6_1_relu[0][0]        
__________________________________________________________________________________________________
conv5_block6_concat (Concatenat (None, 8, 8, 704)    0           conv5_block5_concat[0][0]        
                                                                 conv5_block6_2_conv[0][0]        
__________________________________________________________________________________________________
conv5_block7_0_bn (BatchNormali (None, 8, 8, 704)    2816        conv5_block6_concat[0][0]        
__________________________________________________________________________________________________
conv5_block7_0_relu (Activation (None, 8, 8, 704)    0           conv5_block7_0_bn[0][0]          
__________________________________________________________________________________________________
conv5_block7_1_conv (Conv2D)    (None, 8, 8, 128)    90112       conv5_block7_0_relu[0][0]        
__________________________________________________________________________________________________
conv5_block7_1_bn (BatchNormali (None, 8, 8, 128)    512         conv5_block7_1_conv[0][0]        
__________________________________________________________________________________________________
conv5_block7_1_relu (Activation (None, 8, 8, 128)    0           conv5_block7_1_bn[0][0]          
__________________________________________________________________________________________________
conv5_block7_2_conv (Conv2D)    (None, 8, 8, 32)     36864       conv5_block7_1_relu[0][0]        
__________________________________________________________________________________________________
conv5_block7_concat (Concatenat (None, 8, 8, 736)    0           conv5_block6_concat[0][0]        
                                                                 conv5_block7_2_conv[0][0]        
__________________________________________________________________________________________________
conv5_block8_0_bn (BatchNormali (None, 8, 8, 736)    2944        conv5_block7_concat[0][0]        
__________________________________________________________________________________________________
conv5_block8_0_relu (Activation (None, 8, 8, 736)    0           conv5_block8_0_bn[0][0]          
__________________________________________________________________________________________________
conv5_block8_1_conv (Conv2D)    (None, 8, 8, 128)    94208       conv5_block8_0_relu[0][0]        
__________________________________________________________________________________________________
conv5_block8_1_bn (BatchNormali (None, 8, 8, 128)    512         conv5_block8_1_conv[0][0]        
__________________________________________________________________________________________________
conv5_block8_1_relu (Activation (None, 8, 8, 128)    0           conv5_block8_1_bn[0][0]          
__________________________________________________________________________________________________
conv5_block8_2_conv (Conv2D)    (None, 8, 8, 32)     36864       conv5_block8_1_relu[0][0]        
__________________________________________________________________________________________________
conv5_block8_concat (Concatenat (None, 8, 8, 768)    0           conv5_block7_concat[0][0]        
                                                                 conv5_block8_2_conv[0][0]        
__________________________________________________________________________________________________
conv5_block9_0_bn (BatchNormali (None, 8, 8, 768)    3072        conv5_block8_concat[0][0]        
__________________________________________________________________________________________________
conv5_block9_0_relu (Activation (None, 8, 8, 768)    0           conv5_block9_0_bn[0][0]          
__________________________________________________________________________________________________
conv5_block9_1_conv (Conv2D)    (None, 8, 8, 128)    98304       conv5_block9_0_relu[0][0]        
__________________________________________________________________________________________________
conv5_block9_1_bn (BatchNormali (None, 8, 8, 128)    512         conv5_block9_1_conv[0][0]        
__________________________________________________________________________________________________
conv5_block9_1_relu (Activation (None, 8, 8, 128)    0           conv5_block9_1_bn[0][0]          
__________________________________________________________________________________________________
conv5_block9_2_conv (Conv2D)    (None, 8, 8, 32)     36864       conv5_block9_1_relu[0][0]        
__________________________________________________________________________________________________
conv5_block9_concat (Concatenat (None, 8, 8, 800)    0           conv5_block8_concat[0][0]        
                                                                 conv5_block9_2_conv[0][0]        
__________________________________________________________________________________________________
conv5_block10_0_bn (BatchNormal (None, 8, 8, 800)    3200        conv5_block9_concat[0][0]        
__________________________________________________________________________________________________
conv5_block10_0_relu (Activatio (None, 8, 8, 800)    0           conv5_block10_0_bn[0][0]         
__________________________________________________________________________________________________
conv5_block10_1_conv (Conv2D)   (None, 8, 8, 128)    102400      conv5_block10_0_relu[0][0]       
__________________________________________________________________________________________________
conv5_block10_1_bn (BatchNormal (None, 8, 8, 128)    512         conv5_block10_1_conv[0][0]       
__________________________________________________________________________________________________
conv5_block10_1_relu (Activatio (None, 8, 8, 128)    0           conv5_block10_1_bn[0][0]         
__________________________________________________________________________________________________
conv5_block10_2_conv (Conv2D)   (None, 8, 8, 32)     36864       conv5_block10_1_relu[0][0]       
__________________________________________________________________________________________________
conv5_block10_concat (Concatena (None, 8, 8, 832)    0           conv5_block9_concat[0][0]        
                                                                 conv5_block10_2_conv[0][0]       
__________________________________________________________________________________________________
conv5_block11_0_bn (BatchNormal (None, 8, 8, 832)    3328        conv5_block10_concat[0][0]       
__________________________________________________________________________________________________
conv5_block11_0_relu (Activatio (None, 8, 8, 832)    0           conv5_block11_0_bn[0][0]         
__________________________________________________________________________________________________
conv5_block11_1_conv (Conv2D)   (None, 8, 8, 128)    106496      conv5_block11_0_relu[0][0]       
__________________________________________________________________________________________________
conv5_block11_1_bn (BatchNormal (None, 8, 8, 128)    512         conv5_block11_1_conv[0][0]       
__________________________________________________________________________________________________
conv5_block11_1_relu (Activatio (None, 8, 8, 128)    0           conv5_block11_1_bn[0][0]         
__________________________________________________________________________________________________
conv5_block11_2_conv (Conv2D)   (None, 8, 8, 32)     36864       conv5_block11_1_relu[0][0]       
__________________________________________________________________________________________________
conv5_block11_concat (Concatena (None, 8, 8, 864)    0           conv5_block10_concat[0][0]       
                                                                 conv5_block11_2_conv[0][0]       
__________________________________________________________________________________________________
conv5_block12_0_bn (BatchNormal (None, 8, 8, 864)    3456        conv5_block11_concat[0][0]       
__________________________________________________________________________________________________
conv5_block12_0_relu (Activatio (None, 8, 8, 864)    0           conv5_block12_0_bn[0][0]         
__________________________________________________________________________________________________
conv5_block12_1_conv (Conv2D)   (None, 8, 8, 128)    110592      conv5_block12_0_relu[0][0]       
__________________________________________________________________________________________________
conv5_block12_1_bn (BatchNormal (None, 8, 8, 128)    512         conv5_block12_1_conv[0][0]       
__________________________________________________________________________________________________
conv5_block12_1_relu (Activatio (None, 8, 8, 128)    0           conv5_block12_1_bn[0][0]         
__________________________________________________________________________________________________
conv5_block12_2_conv (Conv2D)   (None, 8, 8, 32)     36864       conv5_block12_1_relu[0][0]       
__________________________________________________________________________________________________
conv5_block12_concat (Concatena (None, 8, 8, 896)    0           conv5_block11_concat[0][0]       
                                                                 conv5_block12_2_conv[0][0]       
__________________________________________________________________________________________________
conv5_block13_0_bn (BatchNormal (None, 8, 8, 896)    3584        conv5_block12_concat[0][0]       
__________________________________________________________________________________________________
conv5_block13_0_relu (Activatio (None, 8, 8, 896)    0           conv5_block13_0_bn[0][0]         
__________________________________________________________________________________________________
conv5_block13_1_conv (Conv2D)   (None, 8, 8, 128)    114688      conv5_block13_0_relu[0][0]       
__________________________________________________________________________________________________
conv5_block13_1_bn (BatchNormal (None, 8, 8, 128)    512         conv5_block13_1_conv[0][0]       
__________________________________________________________________________________________________
conv5_block13_1_relu (Activatio (None, 8, 8, 128)    0           conv5_block13_1_bn[0][0]         
__________________________________________________________________________________________________
conv5_block13_2_conv (Conv2D)   (None, 8, 8, 32)     36864       conv5_block13_1_relu[0][0]       
__________________________________________________________________________________________________
conv5_block13_concat (Concatena (None, 8, 8, 928)    0           conv5_block12_concat[0][0]       
                                                                 conv5_block13_2_conv[0][0]       
__________________________________________________________________________________________________
conv5_block14_0_bn (BatchNormal (None, 8, 8, 928)    3712        conv5_block13_concat[0][0]       
__________________________________________________________________________________________________
conv5_block14_0_relu (Activatio (None, 8, 8, 928)    0           conv5_block14_0_bn[0][0]         
__________________________________________________________________________________________________
conv5_block14_1_conv (Conv2D)   (None, 8, 8, 128)    118784      conv5_block14_0_relu[0][0]       
__________________________________________________________________________________________________
conv5_block14_1_bn (BatchNormal (None, 8, 8, 128)    512         conv5_block14_1_conv[0][0]       
__________________________________________________________________________________________________
conv5_block14_1_relu (Activatio (None, 8, 8, 128)    0           conv5_block14_1_bn[0][0]         
__________________________________________________________________________________________________
conv5_block14_2_conv (Conv2D)   (None, 8, 8, 32)     36864       conv5_block14_1_relu[0][0]       
__________________________________________________________________________________________________
conv5_block14_concat (Concatena (None, 8, 8, 960)    0           conv5_block13_concat[0][0]       
                                                                 conv5_block14_2_conv[0][0]       
__________________________________________________________________________________________________
conv5_block15_0_bn (BatchNormal (None, 8, 8, 960)    3840        conv5_block14_concat[0][0]       
__________________________________________________________________________________________________
conv5_block15_0_relu (Activatio (None, 8, 8, 960)    0           conv5_block15_0_bn[0][0]         
__________________________________________________________________________________________________
conv5_block15_1_conv (Conv2D)   (None, 8, 8, 128)    122880      conv5_block15_0_relu[0][0]       
__________________________________________________________________________________________________
conv5_block15_1_bn (BatchNormal (None, 8, 8, 128)    512         conv5_block15_1_conv[0][0]       
__________________________________________________________________________________________________
conv5_block15_1_relu (Activatio (None, 8, 8, 128)    0           conv5_block15_1_bn[0][0]         
__________________________________________________________________________________________________
conv5_block15_2_conv (Conv2D)   (None, 8, 8, 32)     36864       conv5_block15_1_relu[0][0]       
__________________________________________________________________________________________________
conv5_block15_concat (Concatena (None, 8, 8, 992)    0           conv5_block14_concat[0][0]       
                                                                 conv5_block15_2_conv[0][0]       
__________________________________________________________________________________________________
conv5_block16_0_bn (BatchNormal (None, 8, 8, 992)    3968        conv5_block15_concat[0][0]       
__________________________________________________________________________________________________
conv5_block16_0_relu (Activatio (None, 8, 8, 992)    0           conv5_block16_0_bn[0][0]         
__________________________________________________________________________________________________
conv5_block16_1_conv (Conv2D)   (None, 8, 8, 128)    126976      conv5_block16_0_relu[0][0]       
__________________________________________________________________________________________________
conv5_block16_1_bn (BatchNormal (None, 8, 8, 128)    512         conv5_block16_1_conv[0][0]       
__________________________________________________________________________________________________
conv5_block16_1_relu (Activatio (None, 8, 8, 128)    0           conv5_block16_1_bn[0][0]         
__________________________________________________________________________________________________
conv5_block16_2_conv (Conv2D)   (None, 8, 8, 32)     36864       conv5_block16_1_relu[0][0]       
__________________________________________________________________________________________________
conv5_block16_concat (Concatena (None, 8, 8, 1024)   0           conv5_block15_concat[0][0]       
                                                                 conv5_block16_2_conv[0][0]       
__________________________________________________________________________________________________
bn (BatchNormalization)         (None, 8, 8, 1024)   4096        conv5_block16_concat[0][0]       
__________________________________________________________________________________________________
relu (Activation)               (None, 8, 8, 1024)   0           bn[0][0]                         
__________________________________________________________________________________________________
dense_3 (Dense)                 (None, 8, 8, 1)      1025        relu[0][0]                       
__________________________________________________________________________________________________
up_sampling2d_3 (UpSampling2D)  (None, 256, 256, 1)  0           dense_3[0][0]                    
==================================================================================================
Total params: 7,032,257
Trainable params: 6,948,609
Non-trainable params: 83,648
__________________________________________________________________________________________________

U-Net Model¶

The UNET architecture contains two paths. First path is the contraction path (also called the encoder) which is used to capture the context in the image. The encoder is just a traditional stack of convolutional and max pooling layers.

from tensorflow.keras import Model
from tensorflow.keras.applications.mobilenet import MobileNet, preprocess_input
from tensorflow.keras.callbacks import ModelCheckpoint, EarlyStopping, ReduceLROnPlateau
from tensorflow.keras.layers import Concatenate, Conv2D, UpSampling2D, Reshape
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import ZeroPadding2D, concatenate,Convolution2D, MaxPooling2D, Dropout, Flatten, Activation, BatchNormalization,MaxPool2D,Dense,Input,Concatenate
from tensorflow.keras.utils import Sequence
from tensorflow.keras.optimizers import Adam
from tensorflow.keras.losses import binary_crossentropy
import cv2
from skimage.transform import resize

inputs = Input((128,128,1))
#zeropadding = tf.keras.layers.ZeroPadding2D(padding=1)(inputs)
bn1 = BatchNormalization(momentum=0.9)(inputs)
conv1 = Conv2D(32, (3,3), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(bn1)
conv1 = Conv2D(32, (3,3), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv1)
pool1 = MaxPooling2D(pool_size=(2, 2))(conv1)

bn2 = BatchNormalization(momentum=0.9)(pool1)
conv2 = Conv2D(64, (3,3), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(bn2)
drop12 = Dropout(0.2)(conv2)
conv2 = Conv2D(64, (3,3), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(drop12)
pool2 = MaxPooling2D(pool_size=(2, 2))(conv2)

bn3 = BatchNormalization(momentum=0.9)(pool2)
conv3 = Conv2D(128, (3,3), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(bn3)
drop13 = Dropout(0.2)(conv3)
conv3 = Conv2D(128, (3,3), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(drop13)
pool3 = MaxPooling2D(pool_size=(2, 2))(conv3)

bn4 = BatchNormalization(momentum=0.9)(pool3)
conv4 = Conv2D(256, (3,3), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(bn4)
conv4 = Conv2D(256, (3,3), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv4)
drop4 = Dropout(0.5)(conv4)
pool4 = MaxPooling2D(pool_size=(2, 2))(drop4)

bn5 = BatchNormalization(momentum=0.9)(pool4)
conv5 = Conv2D(512, (3,3), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(bn5)
conv5 = Conv2D(512, (3,3), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv5)
drop5 = Dropout(0.5)(conv5)

bn6 = BatchNormalization(momentum=0.9)(drop5)
up6 = Conv2D(256, (2,2), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(UpSampling2D(size = (2,2))(bn6))
merge6 = concatenate([drop4,up6], axis = 3)
conv6 = Conv2D(256, (3,3), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(merge6)
conv6 = Conv2D(256, (3,3), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv6)

bn7 = BatchNormalization(momentum=0.0)(conv6)
up7 = Conv2D(128, (3,3), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(UpSampling2D(size = (2,2))(bn7))
merge7 = concatenate([conv3,up7], axis = 3)
conv7 = Conv2D(128, (3,3), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(merge7)
conv7 = Conv2D(128, (3,3), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv7)

bn9 = BatchNormalization(momentum=0.0)(conv7)
up8 = Conv2D(64, 2, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(UpSampling2D(size = (2,2))(bn9))
merge8 = concatenate([conv2,up8], axis = 3)
conv8 = Conv2D(64, (3,3), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(merge8)
conv8 = Conv2D(64, (3,3), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv8)

bn8 = BatchNormalization(momentum=0.0)(conv8)
up9 = Conv2D(32, 2, activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(UpSampling2D(size = (2,2))(bn8))
merge9 = concatenate([conv1,up9], axis = 3)
conv9 = Conv2D(32, (3,3), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(merge9)
conv9 = Conv2D(32, (3,3), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv9)
conv9 = Conv2D(2, (3,3), activation = 'relu', padding = 'same', kernel_initializer = 'he_normal')(conv9)

conv10 = Conv2D(1,(1,1), activation = 'sigmoid')(conv9)

model_unet = Model(inputs=[inputs],outputs=[conv10])
optimizer1 = Adam(lr=0.0001, beta_1=0.9, beta_2=0.999, epsilon=None, decay=0.0, amsgrad=False)
model_unet.compile(optimizer = Adam(lr = 0.0001), loss = iou_bce_loss, metrics = ['accuracy',mean_iou])
model_unet.summary()

Model: "model"
__________________________________________________________________________________________________
Layer (type)                    Output Shape         Param #     Connected to                     
==================================================================================================
input_1 (InputLayer)            [(None, 128, 128, 1) 0                                            
__________________________________________________________________________________________________
batch_normalization (BatchNorma (None, 128, 128, 1)  4           input_1[0][0]                    
__________________________________________________________________________________________________
conv2d (Conv2D)                 (None, 128, 128, 32) 320         batch_normalization[0][0]        
__________________________________________________________________________________________________
conv2d_1 (Conv2D)               (None, 128, 128, 32) 9248        conv2d[0][0]                     
__________________________________________________________________________________________________
max_pooling2d (MaxPooling2D)    (None, 64, 64, 32)   0           conv2d_1[0][0]                   
__________________________________________________________________________________________________
batch_normalization_1 (BatchNor (None, 64, 64, 32)   128         max_pooling2d[0][0]              
__________________________________________________________________________________________________
conv2d_2 (Conv2D)               (None, 64, 64, 64)   18496       batch_normalization_1[0][0]      
__________________________________________________________________________________________________
dropout (Dropout)               (None, 64, 64, 64)   0           conv2d_2[0][0]                   
__________________________________________________________________________________________________
conv2d_3 (Conv2D)               (None, 64, 64, 64)   36928       dropout[0][0]                    
__________________________________________________________________________________________________
max_pooling2d_1 (MaxPooling2D)  (None, 32, 32, 64)   0           conv2d_3[0][0]                   
__________________________________________________________________________________________________
batch_normalization_2 (BatchNor (None, 32, 32, 64)   256         max_pooling2d_1[0][0]            
__________________________________________________________________________________________________
conv2d_4 (Conv2D)               (None, 32, 32, 128)  73856       batch_normalization_2[0][0]      
__________________________________________________________________________________________________
dropout_1 (Dropout)             (None, 32, 32, 128)  0           conv2d_4[0][0]                   
__________________________________________________________________________________________________
conv2d_5 (Conv2D)               (None, 32, 32, 128)  147584      dropout_1[0][0]                  
__________________________________________________________________________________________________
max_pooling2d_2 (MaxPooling2D)  (None, 16, 16, 128)  0           conv2d_5[0][0]                   
__________________________________________________________________________________________________
batch_normalization_3 (BatchNor (None, 16, 16, 128)  512         max_pooling2d_2[0][0]            
__________________________________________________________________________________________________
conv2d_6 (Conv2D)               (None, 16, 16, 256)  295168      batch_normalization_3[0][0]      
__________________________________________________________________________________________________
conv2d_7 (Conv2D)               (None, 16, 16, 256)  590080      conv2d_6[0][0]                   
__________________________________________________________________________________________________
dropout_2 (Dropout)             (None, 16, 16, 256)  0           conv2d_7[0][0]                   
__________________________________________________________________________________________________
max_pooling2d_3 (MaxPooling2D)  (None, 8, 8, 256)    0           dropout_2[0][0]                  
__________________________________________________________________________________________________
batch_normalization_4 (BatchNor (None, 8, 8, 256)    1024        max_pooling2d_3[0][0]            
__________________________________________________________________________________________________
conv2d_8 (Conv2D)               (None, 8, 8, 512)    1180160     batch_normalization_4[0][0]      
__________________________________________________________________________________________________
conv2d_9 (Conv2D)               (None, 8, 8, 512)    2359808     conv2d_8[0][0]                   
__________________________________________________________________________________________________
dropout_3 (Dropout)             (None, 8, 8, 512)    0           conv2d_9[0][0]                   
__________________________________________________________________________________________________
batch_normalization_5 (BatchNor (None, 8, 8, 512)    2048        dropout_3[0][0]                  
__________________________________________________________________________________________________
up_sampling2d (UpSampling2D)    (None, 16, 16, 512)  0           batch_normalization_5[0][0]      
__________________________________________________________________________________________________
conv2d_10 (Conv2D)              (None, 16, 16, 256)  524544      up_sampling2d[0][0]              
__________________________________________________________________________________________________
concatenate (Concatenate)       (None, 16, 16, 512)  0           dropout_2[0][0]                  
                                                                 conv2d_10[0][0]                  
__________________________________________________________________________________________________
conv2d_11 (Conv2D)              (None, 16, 16, 256)  1179904     concatenate[0][0]                
__________________________________________________________________________________________________
conv2d_12 (Conv2D)              (None, 16, 16, 256)  590080      conv2d_11[0][0]                  
__________________________________________________________________________________________________
batch_normalization_6 (BatchNor (None, 16, 16, 256)  1024        conv2d_12[0][0]                  
__________________________________________________________________________________________________
up_sampling2d_1 (UpSampling2D)  (None, 32, 32, 256)  0           batch_normalization_6[0][0]      
__________________________________________________________________________________________________
conv2d_13 (Conv2D)              (None, 32, 32, 128)  295040      up_sampling2d_1[0][0]            
__________________________________________________________________________________________________
concatenate_1 (Concatenate)     (None, 32, 32, 256)  0           conv2d_5[0][0]                   
                                                                 conv2d_13[0][0]                  
__________________________________________________________________________________________________
conv2d_14 (Conv2D)              (None, 32, 32, 128)  295040      concatenate_1[0][0]              
__________________________________________________________________________________________________
conv2d_15 (Conv2D)              (None, 32, 32, 128)  147584      conv2d_14[0][0]                  
__________________________________________________________________________________________________
batch_normalization_7 (BatchNor (None, 32, 32, 128)  512         conv2d_15[0][0]                  
__________________________________________________________________________________________________
up_sampling2d_2 (UpSampling2D)  (None, 64, 64, 128)  0           batch_normalization_7[0][0]      
__________________________________________________________________________________________________
conv2d_16 (Conv2D)              (None, 64, 64, 64)   32832       up_sampling2d_2[0][0]            
__________________________________________________________________________________________________
concatenate_2 (Concatenate)     (None, 64, 64, 128)  0           conv2d_3[0][0]                   
                                                                 conv2d_16[0][0]                  
__________________________________________________________________________________________________
conv2d_17 (Conv2D)              (None, 64, 64, 64)   73792       concatenate_2[0][0]              
__________________________________________________________________________________________________
conv2d_18 (Conv2D)              (None, 64, 64, 64)   36928       conv2d_17[0][0]                  
__________________________________________________________________________________________________
batch_normalization_8 (BatchNor (None, 64, 64, 64)   256         conv2d_18[0][0]                  
__________________________________________________________________________________________________
up_sampling2d_3 (UpSampling2D)  (None, 128, 128, 64) 0           batch_normalization_8[0][0]      
__________________________________________________________________________________________________
conv2d_19 (Conv2D)              (None, 128, 128, 32) 8224        up_sampling2d_3[0][0]            
__________________________________________________________________________________________________
concatenate_3 (Concatenate)     (None, 128, 128, 64) 0           conv2d_1[0][0]                   
                                                                 conv2d_19[0][0]                  
__________________________________________________________________________________________________
conv2d_20 (Conv2D)              (None, 128, 128, 32) 18464       concatenate_3[0][0]              
__________________________________________________________________________________________________
conv2d_21 (Conv2D)              (None, 128, 128, 32) 9248        conv2d_20[0][0]                  
__________________________________________________________________________________________________
conv2d_22 (Conv2D)              (None, 128, 128, 2)  578         conv2d_21[0][0]                  
__________________________________________________________________________________________________
conv2d_23 (Conv2D)              (None, 128, 128, 1)  3           conv2d_22[0][0]                  
==================================================================================================
Total params: 7,929,673
Trainable params: 7,926,791
Non-trainable params: 2,882
__________________________________________________________________________________________________

folder = '/content/drive/My Drive/stage_2_train_images'
train_gen = generator(folder, train_filenames, pneumonia_locations, batch_size=42, image_size=128, shuffle=True, augment=True, predict=False)
valid_gen = generator(folder, valid_filenames, pneumonia_locations, batch_size=42, image_size=128, shuffle=False, predict=False)

model_unet.load_weights('/content/drive/My Drive/model_unet_weights6.h5')
history3 = model_unet.fit_generator(train_gen, validation_data=valid_gen, callbacks=[learning_rate], epochs=2)
model_unet.save_weights('/content/drive/My Drive/model_unet_weights8.h5')

Epoch 1/2
57/57 [==============================] - 1916s 34s/step - loss: 0.2912 - accuracy: 0.9595 - mean_iou: 0.7167 - val_loss: 0.2922 - val_accuracy: 0.9559 - val_mean_iou: 0.6721 - lr: 0.0010
Epoch 2/2
57/57 [==============================] - 1832s 32s/step - loss: 0.2849 - accuracy: 0.9608 - mean_iou: 0.7123 - val_loss: 0.2899 - val_accuracy: 0.9571 - val_mean_iou: 0.6862 - lr: 9.9606e-04

model_unet.load_weights('/content/drive/My Drive/model_unet_weights6.h5')
history3 = model_unet.fit_generator(train_gen, validation_data=valid_gen, callbacks=[learning_rate], epochs=4)
model_unet.save_weights('/content/drive/My Drive/model_unet_weights8.h5')

57/57 [==============================] - 1953s 34s/step - loss: 0.2996 - accuracy: 0.9586 - mean_iou: 0.7001 - val_loss: 0.2858 - val_accuracy: 0.9559 - val_mean_iou: 0.7037 - lr: 0.0010
Epoch 2/4
57/57 [==============================] - 1845s 32s/step - loss: 0.2758 - accuracy: 0.9628 - mean_iou: 0.7289 - val_loss: 0.2849 - val_accuracy: 0.9538 - val_mean_iou: 0.7080 - lr: 9.9606e-04
Epoch 3/4
43/57 [=====================>........] - ETA: 7:05 - loss: 0.2744 - accuracy: 0.9637 - mean_iou: 0.7309

plt.figure(figsize=(12,4))
plt.subplot(131)
plt.plot(history3.epoch, history3.history["loss"], label="Train loss")
plt.plot(history3.epoch, history3.history["val_loss"], label="Valid loss")
plt.legend()
plt.subplot(132)
plt.plot(history3.epoch, history3.history["accuracy"], label="Train accuracy")
plt.plot(history3.epoch, history3.history["val_accuracy"], label="Valid accuracy")
plt.legend()
plt.subplot(133)
plt.plot(history3.epoch, history3.history["mean_iou"], label="Train iou")
plt.plot(history3.epoch, history3.history["val_mean_iou"], label="Valid iou")
plt.legend()
plt.show()

for imgs, msks in valid_gen:
    # predict batch of images
    preds = model_unet.predict(imgs)
    # create figure
    f, axarr = plt.subplots(4, 8, figsize=(20,15))
    axarr = axarr.ravel()
    axidx = 0
    # loop through batch
    for img, msk, pred in zip(imgs, msks, preds):
        # plot image
        axarr[axidx].imshow(img[:, :, 0])
        # threshold true mask
        comp = msk[:, :, 0] > 0.5
        # apply connected components
        comp = measure.label(comp)
        # apply bounding boxes
        predictionString = ''
        for region in measure.regionprops(comp):
            # retrieve x, y, height and width
            y, x, y2, x2 = region.bbox
            height = y2 - y
            width = x2 - x
            axarr[axidx].add_patch(patches.Rectangle((x,y),width,height,linewidth=2,edgecolor='b',facecolor='none'))
        # threshold predicted mask
        comp = pred[:, :, 0] > 0.5
        # apply connected components
        comp = measure.label(comp)
        # apply bounding boxes
        predictionString = ''
        for region in measure.regionprops(comp):
            # retrieve x, y, height and width
            y, x, y2, x2 = region.bbox
            height = y2 - y
            width = x2 - x
            conf = np.mean(pred[y:y+height, x:x+width])
            if conf>0.5:
              axarr[axidx].add_patch(patches.Rectangle((x,y),width,height,linewidth=2,edgecolor='r',facecolor='none'))
        axidx += 1
        if axidx==32:
          break
    plt.show()
    # only plot one batch
    break

folder = 'stage_2_test_images'
test_filenames = os.listdir(folder)
# create test generator with predict flag set to True
test_gen = generator(folder, test_filenames, None, batch_size=2, image_size=128, shuffle=False, predict=True)

print('n test samples:', len(test_filenames))
f, axarr = plt.subplots(3, 5, figsize=(20,10))
axarr = axarr.ravel()
axidx = 0
# create submission dictionary
submission_dict = {}
# loop through testset
for imgs, filenames in test_gen:
    # predict batch of images
    preds = model_unet.predict(imgs)
    #axarr[axidx].imshow(resize(imgs[:, :, 0],(128, 128),mode='reflect'))
    # loop through batch
    for img,pred, filename in zip(imgs,preds, filenames):
        # resize predicted mask
        #pred = resize(pred, (1024, 1024), mode='reflect')
        axarr[axidx].imshow(img[:, :, 0])
        # threshold predicted mask
        comp = pred[:, :, 0] > 0.7
        # apply connected components
        comp = measure.label(comp)
        # apply bounding boxes
        predictionString = ''
        for region in measure.regionprops(comp):
            # retrieve x, y, height and width
            y, x, y2, x2 = region.bbox
            height = y2 - y
            width = x2 - x
            # proxy for confidence score
            conf = np.mean(pred[y:y+height, x:x+width])
            # add to predictionString
            if conf>0.76:
              predictionString += str(conf) + ' ' + str(x) + ' ' + str(y) + ' ' + str(width) + ' ' + str(height) + ' '
              axarr[axidx].add_patch(patches.Rectangle((x,y),width,height,linewidth=2,edgecolor='r',facecolor='none'))
        # add filename and predictionString to dictionary
        filename = filename.split('.')[0]
        submission_dict[filename] = predictionString
    #print("--------------------------------------------------------------------------")
    #print(axidx)
    #print(predictionString)
    axidx += 1
    if axidx >= 15: #len(test_filenames)
        break
plt.show()

n test samples: 3000

Model Performance :¶

In this project , implemented three different models (ResNet,U-Net and Dense Net) to identify the pneumonia in lungs.These all three models evaluated with different hyper parameters and different sets of images. Initially we started the evaluation with below 1000 images and kept on increasing the images and captured the Model Accuracy and validation accuracy.For model evaluation ,we have used Jaccord index and Mean IOU technique.In Initial stage , we have achieved the accuracy of 94%,92% and 96% using Resnet,Dense and Unet respectively. To further improve the performance of models, we performed parameter tuning like modifying learning rate, batch size , dropout and batch normalization techniques also increased the network size and optimizer as adam. After Parameter tuning , we have achieved the accuracy of 94%,92% and 96% using Resnet,Dense and Unet respectively. All three models are performing well , compared to others U-Net models have outperformed.

	patientId	x	y	width	height	Target
0	0004cfab-14fd-4e49-80ba-63a80b6bddd6	NaN	NaN	NaN	NaN	0
1	00313ee0-9eaa-42f4-b0ab-c148ed3241cd	NaN	NaN	NaN	NaN	0
2	00322d4d-1c29-4943-afc9-b6754be640eb	NaN	NaN	NaN	NaN	0
3	003d8fa0-6bf1-40ed-b54c-ac657f8495c5	NaN	NaN	NaN	NaN	0
4	00436515-870c-4b36-a041-de91049b9ab4	264.0	152.0	213.0	379.0	1

	patientId	x	y	width	height	Target	class
0	0004cfab-14fd-4e49-80ba-63a80b6bddd6	NaN	NaN	NaN	NaN	0	No Lung Opacity / Not Normal
1	00313ee0-9eaa-42f4-b0ab-c148ed3241cd	NaN	NaN	NaN	NaN	0	No Lung Opacity / Not Normal
2	00322d4d-1c29-4943-afc9-b6754be640eb	NaN	NaN	NaN	NaN	0	No Lung Opacity / Not Normal
3	003d8fa0-6bf1-40ed-b54c-ac657f8495c5	NaN	NaN	NaN	NaN	0	Normal
4	00436515-870c-4b36-a041-de91049b9ab4	264.0	152.0	213.0	379.0	1	Lung Opacity

	patientId	x	y	width	height	Target	class	filenames
0	0004cfab-14fd-4e49-80ba-63a80b6bddd6	NaN	NaN	NaN	NaN	0	No Lung Opacity / Not Normal	0004cfab-14fd-4e49-80ba-63a80b6bddd6.dcm
1	00313ee0-9eaa-42f4-b0ab-c148ed3241cd	NaN	NaN	NaN	NaN	0	No Lung Opacity / Not Normal	00313ee0-9eaa-42f4-b0ab-c148ed3241cd.dcm
2	00322d4d-1c29-4943-afc9-b6754be640eb	NaN	NaN	NaN	NaN	0	No Lung Opacity / Not Normal	00322d4d-1c29-4943-afc9-b6754be640eb.dcm
3	003d8fa0-6bf1-40ed-b54c-ac657f8495c5	NaN	NaN	NaN	NaN	0	Normal	003d8fa0-6bf1-40ed-b54c-ac657f8495c5.dcm
4	00436515-870c-4b36-a041-de91049b9ab4	264.0	152.0	213.0	379.0	1	Lung Opacity	00436515-870c-4b36-a041-de91049b9ab4.dcm

	patientId	x	y	width	height	Target	class	filenames
4	00436515-870c-4b36-a041-de91049b9ab4	264.0	152.0	213.0	379.0	1	Lung Opacity	00436515-870c-4b36-a041-de91049b9ab4.dcm
5	00436515-870c-4b36-a041-de91049b9ab4	264.0	152.0	213.0	379.0	1	Lung Opacity	00436515-870c-4b36-a041-de91049b9ab4.dcm
6	00436515-870c-4b36-a041-de91049b9ab4	562.0	152.0	256.0	453.0	1	Lung Opacity	00436515-870c-4b36-a041-de91049b9ab4.dcm
7	00436515-870c-4b36-a041-de91049b9ab4	562.0	152.0	256.0	453.0	1	Lung Opacity	00436515-870c-4b36-a041-de91049b9ab4.dcm
10	00704310-78a8-4b38-8475-49f4573b2dbb	323.0	577.0	160.0	104.0	1	Lung Opacity	00704310-78a8-4b38-8475-49f4573b2dbb.dcm

	height	width	y	x	Target	patientId
Total	20672.0	20672.0	20672.0	20672.0	0.0	0.0
Percent	100.0	100.0	100.0	100.0	0.0	0.0